Descriptions of New Serpulid Polychaetes from the Kimberleys of Australia and Discussion of Australian and Indo-West Pacific Species of Spirobranchus and Superficially Similar Taxa

In 1988 Pat Hutchings of the Australian Museum, Sydney, undertook an extensive polychaete collection trip off the Kimberley coast of Western Australia, where such a survey had not been conducted since Augener’s (1914) description of some polychaetes from the region. Serpulids were well represented in the collection, and their present study revealed the existence of two new genera, and new species belonging to the genera Protula, Vermiliopsis, Hydroides, Serpula and Spirobranchus. The synonymy of the difficult genera Spirobranchus, Pomatoceros and Pomatoleios is also dealt with. Certain difficult taxa currently referred to as “species complexes” or “species groups” are discussed. For this purpose it was considered necessary to undertake a comparison of apparently similar species, especially of Spirobranchus, from other locations in Australia and the Indo-West Pacific region. It revealed the existence of many more new species, which are also described and discussed below. Pillai, T. GoTTfried, 2009. Descriptions of new serpulid polychaetes from the Kimberleys of Australia and discussion of Australian and Indo-West Pacific species of Spirobranchus and superficially similar taxa. Records of the Australian Museum 61(2): 93–199.


Material and methods
The islands off the Kimberley Plateau, North Western Australia, from where the serpulids described in this paper were collected are shown in Figure1.
All available specimens, from juveniles to adults were studied in order to obtain as much data as possible on ontogenetic changes in characters, such as, in the operculum, number and arrangement of radioles, abdominal segments and arrangement of abdominal uncinal tori.
Many specimens possess complete tubes, some only tube fragments, while the tubes of others are missing. Since tube characteristics are best observed prior to extracting the worms from within, the following procedures were carried out under a low-power stereomicroscope with the specimens placed under alcohol in a laboratory dish. They were first cleaned using a paintbrush with stiff bristles. Calcareous overgrowths were removed with a fine-tipped surgical scalpel avoiding damage to tube ornamentation, and then measured and illustrated. To extract the worms, each tube was fractured as follows. It was firmly held with a pair of forceps; then the blunt edge of an ordinary laboratory scalpel was placed across it with its pointed end resting on the dish, and appropriate pressure was applied at intervals, from its anterior to posterior end, taking care not to damage the worm within. For juvenile specimens, a firm bent mounted needle was used in a similar manner, applying pressure along the tube with the convex side of its bend.
The following routine measurements were carried out, depending on completeness of available specimens. Tubes: total length, maximum diameter; height at aperture and maximum width, including lateral flanges. Worms: total length from tip of longest radiole or operculum, whichever is longer, to posterior end of abdomen; width of thorax at level of collar fascicles; length of thorax from collar to last thoracic segment; length of operculum plus peduncle; length and diameter of opercular plate; longest radiole, including, if found uncontracted, its pinnule-free tip; and length of abdomen. Counts: number of radioles on each side, and, if spirally arranged, the number of spirals to record maximum number encountered; number of pairs of thoracic chaetigers, or number on each side, if different; number of abdominal segments, and number bearing capillaries at its posterior end.
Slides were prepared in Gurr's Aquamount of the following: collar chaetae teased out form an entire collar fascicle, some thoracic and anterior abdominal tori, and the terminal abdominal segments of one side. Carbon-dioxide bubbles released by the mounting medium reacting with calcareous matter in the tissues were lifted off with the bend of a curved mounted needle before placing the cover glass, and then baked in a slide oven.
Drawings were made under low and high power magnifications with the aid of drawing attachments. For collar chaetae, they consisted of the blade, the boss and part of the chaetal shaft. For detailed comparisons between species, they included those recently formed within the collar fascicles and had not been subjected to abrasion, as well those earlier formed ones projecting outside the fascicle that had been subjected to abrasion with use. The number and arrangement of teeth in the thoracic and abdominal uncini and shape of their most anterior tooth were noted.
However, research on serpulids carried out by numerous scientists during the period of over eighty years since then, had resulted in accrual of a vast amount of knowledge on them and alternative schemes for their classification, and there appear to be no sound reasons for reverting to Fauvel's (1927) classification rather than moving on, on the basis of knowledge available since then.
An important morphological character that needs to be taken into consideration with regard to classification of serpulids in general is that of opercular insertion. Although ten Hove (1984: 184-189) did not categorically state that the operculum in all serpulimorphs is derived from the second dorsalmost branchial radiole, his arguments regarding it in certain taxa, supported by drawings, could, perhaps, lead one to such a conclusion. Indeed, Fitzhugh (1989: 11) concludes that "ten  has shown that the serpulid operculum is actually a modified second, dorsalmost radiole. It is only in large-bodied serpulid species that the opercular stalk migrates during development to a position, which causes it to appear as though it originates from the first radiole." Rouse (2000: 187) came to the same conclusion: "ten  argued that in all serpulids the operculum is derived from the second most dorsal radiole, though its position may alter subsequently." However, this is, in reality, not the case with with all serpulids, as shown by Matjjašic & Sket (1966), Orrhage (1980) and Pillai (2008Pillai ( , 2009. In the latter context, Fauvel's inclusion of Protula under the Filograninae causes some problems. According to Kupriyanova & Jirkov (1997), in Protula, an operculum, when present, is vesicular and occurs in the position of the second branchial radiole, whereas Filograna has "two spoonshaped opercula on the first non-modified pinnulate branchial radioles, one on each branchial radiole". Likewise, Fauvel's subfamily Serpulinae includes genera, such as, Serpula and Spirobranchus in which the operculum is inserted in the position of the second radiole as well as those, such as, Ditrupa, Vermiliopsis, Pomatostegus and Mercierella (currently Ficopomatus) in which it inserted in the position of the first radiole. Uchida (1978) erects several new serpulid subfamilies, the acceptance of which has been met with some reticence. However, it is to be expected that adequate scientific reasoning be provided in the event of acceptance or nonacceptance any of his specific proposals. In the meantime, classification of the Family Serpulidae into subfamilies in this paper is limited to a few that could adequately be explained and accepted in terms of the International Rules for Zoological Nomenclature. In the latter context, the subfamily Serpulinae, Macleay, 1840, sensu stricto is used for genera that constitute the Serpula-Spiraserpula-Crucigera-Hydroides clade.
The descriptions of taxa in this paper broadly follow current phylogenetically based serpulimorph classifications. Their groupings range from genera that are currently believed by various authors to be more primitive, e.g., the Filograninae, to those less primitive and possess either simple to complex chitinous opercula, or calcareous opercula, besides other characters.
They are dealt with alphabetically in two sections: the Kimberleys genera and species, followed by comparisons with species, especially of Spirobranchus, from other locations in Australia and the Indo-West Pacific. Within each genus, the species are dealt with in alphabetical order.
The present paper follows ten  who, based on the literature, states that the tropical Indo-West Pacific records can be roughly brought together under two taxa."One of them is Protula bispiralis with 3-8 whorls of spirally arranged branchial radioles and a tube diameter of up to two centimeters. The second is Protulopsis nigranucha (Fischli, 1900) which has "a semicircular arrangement of its radioles (at most in a ¾ spire), and with a tube of up to a few millimetres in diameter." It is evident that ontogenetic studies relating to the following characters would be useful in the revision of the genus: the presence or absence of spirally arranged radioles and maximum number of spirals in relation to size, presence or absence of an apron, and whether the latter is present in juveniles and lost in adults.

Description
Measurements. Maximum external diameters of two largest tubes, respectively: 13.0 mm and 12.6 mm, their thicknesses, 1.5 mm and 1.3 mm, total lengths of two worms from AM W21439: 80.0 mm and 75.0 mm; their thoracic widths of 9.5 mm and 8.6 mm, respectively. Abdominal segments in two larger specimens number 155 and 159, with capillaries occurring from about 81st to 109th and 81st segments, respectively, until about 4 segments from their posterior ends.
Remarks. For a good species description, see Day (1967: 818). Accordingly, it attains very large sizes, up to a length of 180.0 mm, and there may be up to 10 whorls of radioles on each side arranged along a central axis, united for a third of their length by a web (= inter-radiolar membranes). The largest specimen in the Kimberleys collection is less than half that size; it has 4 radiolar whorls on each side. However, NHM collection 1905BMNH 1924.181a contains very large specimen from Cargados Carajos, NE of Mauritius, Indian Ocean, collected by J.S. Gardiner, and identified as P. bispiralis by ten Hove. Although the worm is contracted in alcohol, and the radioles difficult to measure, the total length of the worm is well over 150.0 mm; its thorax is 60.0 mm long and 25.0 mm wide; its abdomen 05.0 mm long and its maximum width of 20. 0 mm; and nine radiolar whorls could be counted on each side. A specimen from South Africa, in NHM collection BMNH 1936.10.16.312, also has 9 whorls on each side.
The thoracic uncinal tori are in P. bispiralis apparently lost as the worms grow older. A similar situation also apparently prevails in P. tubularia from the Mediterranean, in which, according to Fauvel (1927: 381), their presence is not constant, occurring most frequently in smaller individuals and apparently totally lacking in larger ones. The thoracic membranes extend to the last thoracic chaetigers, but do not form an apron. Ontogenetic studies would help to confirm whether in P. bispiralis too an apron is present in juveniles and lost in specimens. Meanwhile, following ten , the above specimens are assigned to Protula bispiralis.
Remarks. All taxa presently considered to belonging to the "Protula/Apomatus complex" possess an apron. Some may or may not possess an operculum. Pseudoprotula lacks an apron.

Description
Measurements. Holotype (complete worm) total length 10.3 mm; width of thorax: 1.2 mm; length of abdomen 5.9 mm. Paratype: incomplete, lacks radioles; length without radioles about 15.5 mm; width of thorax 1.2; length of abdomen 10.5 mm. Holotype (Fig. 2H,I) with 11 radioles on left and 14 on right, pinnule-free tips short; its abdomen has 79 segments, with capillaries on about the last 30 segments. Paratype has 93 abdominal segments, with capillaries from about the 77th.
Tube. Missing in holotype and paratype.
Worm. Operculum absent; thorax with 7 pairs of chaetal fascicles; thoracic membranes broad and end on 5th chaetiger (Fig. 2I); first five thoracic uncinal tori on each side borne along posterior edge of a wide flap-like extension, posterolateral to its corresponding thoracic notochaetal fascicle; last uncinal torus on each side borne on a similar, larger, flap-like expansion. Corresponding flap-like expansions of the two sides not joined to form an apron (Fig. 2H).
Etymology. Named after the type locality, off the Kimberleys.

Description
Measurements. External tube diameter: from 4.3 mm in an adult specimen in AM W21471 to 0.6 mm in an early juvenile (Fig. 4D). Total lengths of worms: 25.3 mm in a specimen from AM W21471 to 6.5 mm in smallest juvenile. Thoracic widths: 2.5 mm to 0.5 mm in smallest available specimen in AM W21471. Length of operculum plus peduncle: 6.0 mm to 1.5 mm in smallest specimen. Length of operculum alone: 0.5 mm in smallest specimen to 2.5 mm.
Tube. White, mostly covered over by coral encrustation; difficult to remove without damage to ornamentation. Holotype (Fig. 3A) bears 5 longitudinal ridges in earlier formed posterior part, up to 7 anteriorly, where it is trapezoidal in cross-section. A juvenile paratype also bears 5 longitudinal ridges, although roughly circular in crosssection (Fig. 4A).
Worm. Radioles arranged in a semicircle on each side, their numbers ranging from 8 in a small juvenile to 22 in an older specimen; their pinnule-free tips club-shaped and thicker than pinnules; an early juvenile (Fig. 4E) has very few radioles on each side, somewhat swollen tips, and its operculum not yet developed. Operculum occurs on left or right side; rudimentary operculum absent. It consists of a bulbous proximal part and a chitinous distal part; the latter measuring 2.1 mm in a larger specimen, to 0.7 mm in a small juvenile. Chitinous distal part is characteristically cylindrical rather than conical; quite transparent ( Fig.  3D,I,M); its diameter greater than its height, giving it the appearance of a circular pillbox, even in specimens with with a single partition (   showing its short and broad chitinous operculum, its cap contracted into a concavity, with a central peak, and the extent of the thoracic membranes; (C) left ventrolateral view of thorax showing the ocellar cluster, termination of the thoracic membrane, the contracted proximal bulb-like part of the operculum, the peduncle, and numerous radioles; (D) enlarged drawing of operculum showing its transparent chitinous distal part, a single chamber with its terminal part contracted into a concavity, and peduncle contracted into several irregular folds; (E) the cushion-like swelling (pygidial gland) at the posterior end of abdomen; (F) thoracic uncini; (G) anterior abdominal uncini; (H,I) two views of the operculum of specimen from AM W21434 showing an un-contracted dome-shaped distal part, a single transparent chamber below it, its contracted proximal bulb-shaped part (partly into the transparent chamber), and contracted peduncle; (J) anterior end of a small juvenile specimen showing operculum and extent of thoracic membranes; (K) tube fragment found together with a specimen from AM W21434 (doubtfully included here); confirmation necessary as to whether such a circular unattached part can occur in this species. L,M, specimen from AM W21434: (L) anterior end of worm; (M) operculum; both show an un-contracted proximal bulb-shaped part, and a cylindrical distal part, the latter consisting of 3 partitions preceding the terminal chamber; (N) operculum of a specimen from AM W21471: it is similar to that of the preceding specimen, but has two partitions proximal to the cylindrical chamber. cap slightly convex, dome-shaped; bears a small peak in its centre in specimens unaffected in alcohol preservation (Fig. 3H,I,. Both the terminal cap and peak may be collapsed into to a concavity in those affected by alcohol preservation, (Fig. 3B,D).
Chaetae. Chaetal composition as follows: those of first or collar fascicle similar to those of species described below; 2nd and 3rd thoracic chaetigers consist of bladed chaetae and capillaries; fourth consists of bladed chaetae, capillaries and transitional sickles (i.e., sickles commence on the fourth thoracic chaetiger); and 5th to 7th of bladed chaetae, capillaries and sickles. Each uncinal torus borne along the posterior border of a flap-like posteroventral extension of the thoracic wall (Fig. 3C); thoracic and anterior abdominal uncini saw-shaped (Fig. 3F,G); rasp-shaped in posterior part of abdomen; number of teeth in thoracic uncini 14 or 15; most anterior uncinal tooth simple, with a truncated tip (Fig. 3F,G).
Remarks. Microscopic examination of the early juvenile worm in which an operculum is yet to be developed (Fig. 4E), revealed the presence of sickle-shaped chaetae in the thorax.
The number of peduncular annuli is variable, a maximum number of about 10 is present in one specimen (Fig. 3I); less in another (Fig. 3N); not clearly formed in some (  A circular, detached portion of a tube, having transverse growth rings, was also found with the specimens in AM W21434 (Fig. 3K). Further studies are required to confirm whether a circular unattached anterior part occurs in this species.

Description
Measurements. External tube diameters: 0.5 mm in juveniles to 2.0 mm in older specimens. Total lengths of worms: 3.2 mm in a juvenile to 13.7 mm in a larger specimen; their thoracic widths: 0.7 mm to 0.2 mm, respectively. Lengths of operculum and peduncle range from 3.8 mm in an older specimen to 1.0 mm in smallest juvenile; length of abdomen up to about 7.5 mm, possibly longer in specimens not contracted in alcohol.
Tube. Shape: sinuous (Figs 5A, 6B,K) to loosely coiled (Figs 5C,6 I,7 I). Maximum number of longitudinal ridges observed 5 (Figs 5A,C). When sinuous, outermost ridge on each side, especially on concave side, extends laterally rather than vertically. Earlier formed posterior parts of tubes may have a smaller number of longitudinal ridges, e.g., 4 (Fig. 6B). Texture: mostly more or less pitted in longitudinal rows between ridges; may occasionally be quite smooth and shiny, with smooth ridges and hardly any pitting (Fig. 7 I); or somewhat smooth-ridged and regularly pitted (Fig. 5C); older tubes roughly trapezoidal in cross-section, with a wide lateral flange along the substratum on each side, as seen in two latter specimens; a light brown inner tube lining often present. An unattached tube fragment, circular in crosssection, occurring in collection has peristome-like transverse ridges and some faint transverse ridges between latter (Fig.  7B). Further studies are necessary to confirm whether such an unattached portion occurs in this species.
Worm. Radioles semicircularly arranged; maximum number observed per side: 10 in an older specimen, 3 in smallest juvenile; pinnule-free tips 1 ⁄8 to 1 ⁄9 the total length of the radioles, usually slightly thicker than the pinnules (Figs 5A,B,D,E, 6J-L, 7C-E,G,N); up to about thrice as thick as pinnules in some specimens. Determination of opercular insertion difficult in adult worms; ontogenetically, however, it occurs in position of first, as seen in juveniles (Fig. 7D,E,M,N); may occur on the left or right side; rudimentary operculum absent. In older specimens, consists of a bulbous transparent to translucent basal part, and a conical transversely partitioned distal cap of chitin-like material (Figs 5A,B, 6C-E,I-L, 7A,C-E,G,H); latter translucent to semi-translucent, light to somewhat darker-brown in colour. Number of partitions ranges from none in the smallest juvenile (Fig. 7D, K,M,N), to 9 in one of the older specimens (Fig. 5A); however, maximum number could be much higher, 16, (Fig. 7A). Variations in opercular shape in alcohol-preserved material: as follows. Terminal cap in early juveniles frequently collapsed into a concavity with a small convex peak at its centre K,M,N). Likely, however, to be dome-shaped in live condition, as seen in less contracted juvenile operculum (Fig. 7G,H); may be relatively unaffected (Fig. 5A,B), or collapsed to various extents in older specimens too (Figs 5D,E, 6J-L). Basal bulb-shaped part of operculum may also may retain its original form (Figs 6K,L,7K,M,N), or be affected to various extents (Figs 5A,B,D, 6C-E,7A,C,D,G,H,J). Peduncle bears up to about 5 annuli distally, most distal one conspicuously large; may be relatively unaffected in alcohol preservation (Figs 6J,L,7A,E,G,H,M,N); or, frequently, assume a very characteristic ventrally directed projection (Figs 5E,6C,D).
Thorax bears 7 pairs of chaetal fascicles and 6 pairs of uncinal tori (Figs 5B,E, 6D,J,L, 7J). Thoracic membranes extend to fourth or fifth chaetiger only (Figs 5D, 6D,J,L, 7C-E,G); apron absent. Total number of abdominal segments varies from 32 in the smallest specimen available to 66 in an older specimen; capillaries in latter specimens start from segments 23 and 51, respectively; pygidial gland absent.
Ontogenetic observations. The small juveniles in collection provide data on early development of tube, besides derivation and later development of the operculum and number of branchial radioles. The circular juvenile tube condition is shown in Fig. 7K, which then develops a MLR (Fig. 7L). A LLR is then added on, e.g., as seen toward the posterior end of the tube in Fig. 7I. Then, an OLR is added on each side, bringing the total to 5 as shown in Fig. 7F,I. Their form and number may or may not be distinct along parts of the tube (Fig. 6K).
As seen in the earliest available juvenile, the operculum is shaped like an inverted bell, terminating in a transparent to translucent chitinous plate (Fig. 7K,M,N). A faint constriction is discernible distally, between the vesicular part and the chitinous plate. At the centre of the latter is a tiny transparent to translucent swelling or peak. The first peduncular annulus is developed just beneath the vesicle (Fig. 7M). Two such annuli are seen in a later juvenile (Fig. 7C,D). It also bears a somewhat more prominent peak at the centre of the latter. In the next stage (Figs 6C-E, 7G,H), the operculum has changed shape, its chitinous part becoming somewhat conical, and its distal plate more like a cap, and its peduncle developing more annuli, most of them partially formed. An additional partition is noticeable in the cap (Fig. 7H).
The early juvenile operculum of V. glandigera as described above goes through stages similar to those described for V. cylindrica (Figs 3J,4B,C), and it appears difficult to distinguish between the two species at this stage.
Remarks. The differences between V. cylindrica n.sp., and V. glandigera can be summarized as follows. In V. cylindrica, worms attain a larger size, as observed, up to a total length of about 25.3 mm, and possess a maximum number of 22 branchial radioles, which is about double that in V. glandigera. The chitinous distal part of the operculum in older specimens of V. cylindrica is cylindrical, its diameter greater than its height, which gives it a cylindrical pillbox shape transparent and may bear from 1 to about 5 light-brown bands encircling it.. In V. glandigera, on the other hand, it is conical in shape, and the maximum number of transverse partitions encountered is much higher, up to 16 observed in the Kimberleys collection (Fig. 7A). Thoracic uncini in V. cylindrica bear about 14-15 teeth and a truncated anterior process (Fig. 3H), whereas the number in V. glandigera is somewhat lower, about 12, and its anterior process is pointed (Fig. 6G).  , figured from two specimens from AM W21470; A and B from first specimen; C-E from a second specimen: (A) worm in situ within its tube; (B) worm removed from tube showing the transparent partitioned operculum, its bulb-shaped basal part, the conspicuous annulus just below the latter, pinnule-free radiolar tips and extent of the thoracic membranes; (C) tube; D,E, worm removed from its tube: (D) worm showing distal end of the chitinous part of the operculum contracted into a concavity, and bears a small digitiform process in its centre; (E) another view of worm showing the transparent distal part of the operculum; the large annulus at the distal end of the peduncle has assumed the shape of a ventrally directed projection in alcohol.
However, there have been problems concerning the separation of V. glandigera occurring in the Indo-West Pacific from the Mediterranean and North Atlantic species V. infundibulum (Philippi), 1844. This has been due to recognition of similarities between material originally described from the Red Sea as Vermiliopsis glandigerus (= Vermiliopsis glandigera) by Gravier (1908: 121) and the type species of the genus Vermiliopsis, as stated above, namely, Vermiliopsis multivaricosa Mörch, 1865(= V. infundibulum Philippi, 1844, but not differences in tube structure. Monro (1930: 211) identified material from the Gulf of Guinea, off the Northwest Atlantic coast of Africa, as V. glandigerus, and thought it probable that V. glandigerus Gravier, and a couple of other species "will prove to be simple varieties of V. infundibulum." Subsequently, authors working on Vermiliopsis from the Indo-West Pacific have Figure 6. A-L, Vermiliopsis glandigera  A-E, a juvenile specimen from AM W21470: (A,B) portions of the tube; (C,D) two views of worm removed from tube. The operculum has few partitions, the enlarged peduncular annulus just below the operculum and the extent of the thoracic membrane of the left side are shown; (E) another view of the same operculum. F-H, chaetae of a specimen from AM W21470: (F) collar chaeta; (G) thoracic uncini; (H) anterior abdominal uncini. I,J, a specimen from AM W21451: (I) tube with worm within; (J) worm removed from tube showing the number of partitions in the operculum, the enlarged peduncular annulus just below the operculum, and the extent of the thoracic membrane on the right side. K,L, a specimen from AM W202948: (K), worm and tube in situ; (L), worm removed from tube; the operculum has 12 partitions, the peduncular annulus below the operculum is enlarged. referred the two species to a "Vermiliopsis infundibulum/ glandigera-group," e.g., Imajima (1976: 139;1978: 57;1979: 169), Imajima & Hove (1984: 59, 1986: 9), Ishaq & Mustaquim (1996: 182), Nishi (1996: 315); or to a "Vermiliopsis infundibulum/glandigera-complex", e.g., ten Hove (1994: 113), and Fiege & Sun (1999: 133). Ten Hove (1994: 113) also draws attention to the difficulties in identifying species belonging to the "Vermiliopsis infundibulum/glandigera-complex", in referring to them as being "one of the most difficult complexes of serpulids".  . (A) operculum of a specimen from AM W21470 with 16 partitions, including the terminal cap, and showing the characteristic enlarged terminal annulus of the peduncle; (B) part of a tube found with latter specimen, questionably, an anterior unattached part. C,D, Two juveniles from AM W21472; (C) worm showing distal part of its operculum contracted into a concavity with a central peak, partitions not yet developed, annulus below its basal bulb-like part is already enlarged, and thoracic membranes extend to the 4th chaetiger; (D) anterior end of a juvenile worm from AM W21472: operculum has 2+ partitions, its distal cap-like annulus is contracted into a concavity, and bears a central peak; thoracic membranes extend to end of fourth chaetiger; (E) a specimen from AM W21472 having 3+ opercular partitions, its distal cap contracted into a concavity with a peak, and enlarged annulus at the distal end of its peduncle is discernible; (F-H) a juvenile specimen from AM W21472: operculum has an uncontracted dome-shaped cap with a peak, and three partitions; annulus at the distal end of its peduncle is enlarged. What appears to be a pygidial gland is faintly developed. I-J, a specimen from AM W21470: (I) tube, more regularly formed than in other specimens; smooth and shiny, not strongly pitted, has 3 smooth longitudinal ridges, besides the laterally extended one on each side, and well-formed transverse ridges (peristomes) at intervals. Worm (J) is indistinguishable from other specimens, including the swelling at the distal end of the peduncle. K-N, an early juvenile from AM W21470: (K) worm within part of its tube; (L) tube shows fine transverse ridges and a single irregularly serrated median dorsal ridge; (M) anterior end of worm; (N) whole worm; M,N, opercular cap contracted into a concavity and bears a rounded peak; the swollen distal annulus of the peduncle being developed. Furthermore, based on past records of material identified as V. infundibulum (Philippi, 1844) from the Mediterranean to the Caribbean, Brazil, West and South Africa, Indian and Pacific Ocean by several authors, Zibrowius (1966Zibrowius ( : 376, 1968 considered its distribution to be cosmopolitan. According to other studies, however, most of those earlier records outside the Mediterranean and the North Atlantic have been misidentifications. It would, therefore, appear useful to compare what Monro (1930) identified as V. glandigera from the Gulf of Guinea with the Kimberleys material described above as V. glandigera in order to separate the two species that had been grouped together as the "Vermiliopsis infundibulum / glandigera complex" or "group" in the Indo-West Pacific.
An important character that had hitherto not been adequately recognized pertains to the peristomes occurring along the tube. They are considered here to be quite characteristic of V. infundibulum. Indeed, they are the only structures that could be recognized as being infundibular (funnel-shaped) and incorporated into its specific name by its author, Philippi, 1844. A concise and apt description is provided by Fauvel (1927: 363, fig. 124c): "Tube calcaire, blanchâtre, épais, rapidment évasé, bourrelets en entonnoir." Zibrowius's description (1966: 376/377) of tubes of of V. infundibulum collected from Madeira during the "Jean Charcot" Expedition is similar: "Les tubes … toujours pourvus de péristomes évasés." In a later, more detailed study, Zibrowius (1968a: 122) confirms the presence of funnelshaped peristomes along the tube of V. infundibulum, and adds the following: that it may be quite smooth or possess, frequently 4 or 5, or rarely 2 or 7, longitudinal ridges, but characterized by the "funnel-shaped" peristomes. The terminal peristome is always better developed and more outwardly directed than the preceding ones. A smooth tube with vase-shaped peristomes is figured and described by Bianchi (1981: 72, fig. 25a;see Fig. 8B in present paper). However, the tube fragment figured by Monro (1930: fig. 88c;see Fig. 8C in present paper) fits neither that of V. infundibulum, nor of V. glandigera described from the Kimberleys, but of another species. The terminal annulus of the opercula figured by Monro (1981: fig. 88a) is also different from that of V. glandigera from the Kimberleys.
The more important characters that distinguish V. infundibulum and V. glandigera can therefore be summarized as follows.

Differences between Vermiliopsis infundibulum and Vermiliopsis glandigera
In Vermiliopsis infundibulum, tubes bear characteristic massive funnel-shaped to vase-shaped peristomes with thickened outwardly directed rims; may be smooth or bear a number of longitudinal ridges but bear the characteristic peristomes; terminal peristome always better developed and outwardly directed than preceding ones. Worms attain maximum lengths of up to 30.0 mm; possess up to 20 radioles. Operculum could exceed 8.0 mm, chitinous, brown; more or less transparent; may possess up to 70 partitions. Thoracic uncini may possess up about 16 teeth; most anterior tooth longer than rest, simple. Number of abdominal segments may exceed hundred. See Zibrowius (1968a: 121-123).
In V. glandigera, on the other hand, tubes may be flared outwards on the substratum, but do not possess vase-shaped or funnel-shaped peristomes (Figs 5A,C,6A,B,I,K,7F,I). Indeed, such vase-shaped or funnel-shaped peristomes have not been described for the "V. infundibulum/glandigeragroup" recorded from various localities in Indo-West Pacific. Instead, peristomes, when present, consist of thickened transverse ridges. A number of longitudinal ridges usually present. As regards Fig. 7B, even if such erect terminal portions of tubes are confirmed to occur sometimes in V. glandigera, their peristomes are not comparable to those of V. infundibulum (Fig. 8A,B). Worms: attain lengths of up to about 13.7 mm. Maximum number of radioles observed: 10 on each side. Operculum and peduncle attain lengths up to about 3.8 mm; maximum number of partitions observed in chitinous distal part, 16. Thoracic uncini: bear about 12 teeth and a simple anterior process.Maximum number of abdominal segments observed 66. The maximum attainable size, number of branchial radioles on each side, and number of abdominal segments in V. infundibulum is, therefore, about double that of V. glandigera. The maximum number of opercular partitions observed in former is about four times that of the latter. The number of teeth in the thoracic uncini is also higher. Zibrowius (1968a: 123) gives the distribution of V. infundibulum as the Eastern Atlantic and Mediterranean, it's having been reported from Angola, Azores, Canaries, numerous locations in the Mediterranean, and the Black Sea. The latter distribution, the characters of V. infundibulum discussed above, and Monro's (1930) tube fragment (Fig.  8C) and opercula confirm that Monro's material from the Gulf of Guinea was not V. glandigera. Vermiliopsis infundibulum can now be, therefore, eliminated from the Indo-West Pacific. Past identifications of collections from the Indo-West Pacific as the latter may now be referred to Vermiliopsis glandigera , or other species of Vermiliopsis, as appropriate.
Ten Hove & Kupriyanova (2009: 100) provide an important discussion in this context. They use the name of the type species of genus Vermiliopsis Saint-Joseph, 1894, namely, Vermilia multivaricosa Mörch, 1861, as the new name for Vermilia glandigera sensu Philippi, 1844. Since Vermilia has been replaced by Vermiliopsis, the "Vermiliopsis infundibulum/glandigera-complex" or "group" discussed above would change to "Vermiliopsis multivaricosa/ glandigera-complex" or "group". Ten Hove & Kupriyanova, (2009: 102) refer to V. infundibulum as part of a complex with V. pygidialis and V. glandigera in the Indo-West Pacific. Imajima (1976) describes material from Southwest Japan as "Vermiliopsis infundibulum/glandigera group." It attains nearly twice the size of V. glandigera, the largest worm having a total length of 25 mm and 109 segments and 18-19 pairs of branchial radioles. Moreover, it bears Apomatuschaetae from the third thoracic chaetiger onwards i.e., the 3rd to 7th thoracic chaetigers, and the thoracic uncini bear 15-17 teeth. Quite interestingly, and as specifically described and illustrated by Imajima (fig. 11a,, "several dark brown rings". Such pronounced dark rings or bands were not reported for V. glandigera by , are not obvious in the Kimberleys specimens, and have not been described in other collections identified as the "Vermiliopsis infundibulum  Fauvel (1927: fig. 124c), magnified about 4 times that of Fauvel's figure which represented its natural size. It also shows ontogenetic changes in which the earliest formed part abruptly increases in diameter and develops the vase-shaped peristomes, already bearing 11 of them in the tube of approximately 48.0 mm in length; (B) after Bianchi (1981: fig. 25a), same size as original figure; evidently an older tube compared to A; note the well-formed terminal peristome and thickened edges. (C) After Monro (1930: fig. 88c), same size as original figure; tube fragment figured by Monro as belonging to Vermiliopsis glandigera . Note that none of three types of tubes A-C occur in V. glandigera described from the Kimberleys. /glandigera group" from the Indo-West Pacific. Further work is necessary to determine whether Imajima's specimens from Southwest Japan belong to a hitherto un-named species. Fiege & Sun (1999: 133-136, fig. 21) describe material from the South China Sea as belonging to the "Vermiliopsis infundibulum/glandigera group". Since only anterior fragments were available, or they were retracted into their tubes, the authors did not provide the total length of the worm, numbers of radioles or the extent of the thoracic membranes. The opercula and their peduncles show similarities to those of Gravier's V. glandigera from the Red Sea, as well as the Kimberleys material belonging to the latter. They are also similar in not having sharply demarcated dark brown bands that are characteristic of the opercula of Imajima's species. However, the number of teeth in their thoracic uncini is around 15, whereas it is around 12 in V. glandigera. Apomatus-chaetae commence on the 3rd thoracic chaetiger in the South China Sea specimens, while they commence on the 4th in the Kimberleys specimens of V. glandigera. Data on the South China Sea specimens appear to be inadequate to determine whether they belong to V. glandigera or another species.

Indeterminate species of Vermiliopsis reported from the Indo-West Pacific
The material from Pakistan described by Ishaq & Mustaquim (1996: 182-183) as belonging to the "V. infundibulum/glandigera group" appears to possess some important differences from V. glandigera. The worms attain a total length of up to about 18.0 mm, the number of radioles is 10 to 12 on each side, thoracic membranes reach up to the end of the 4th thoracic chaetiger, Apomatus-chaetae occur from the 4th chaetiger onwards, and thoracic uncini have 10-12 teeth. Its description hitherto is similar to that of V. glandigera. However, there are significant differences. A maximum number of up to 114 abdominal segments have been observed in the material from Pakistan, as opposed to about half that number in V. glandigera. The most important difference is with regard to the operculum of the material from Pakistan, in which transverse partitions are absent even in worms larger than V. glandigera. Furthermore, Ishaq & Mustaquim describe the shape of the operculum as being ovoid to conical and, according their figure (fig. 11A) the dome-shaped horny opercular cap is blackish or black. The species is, evidently, neither V. infundibulum nor V. glandigera.
The genus is presently known to consist only of its type species. With its distribution apparently extending from the tropical western Atlantic to the Indo-Pacific it appears to be an ancient serpulid taxon.
Worm. Operculum on left or right side; peduncle fleshy and winged (Fig. 9E), bears a chitinous tiered distal part (Fig.  10A). Latter consists of a central column bearing circlets of dark outwardly directed spines at intervals, above each of which is a thin, shallow, translucent, yellowish-brown, saucer-shaped structure with a scalloped margin; marginal scallops not regularly shaped, but more or less rounded, flattened and their tips directed posterolaterally. Grooves extend between and along scallops to, or part of the way towards central column. Saucer-shaped structures may completely or partly damaged or lost in earlier formed distal tiers. Maximum number of tiers observed in collection 6. Maximum number of radioles observed per side, 58. Pair of slit-like apertures representing a pair of ventral thoracic glands present, (Fig. 9D). Pair of greyish/blackish pigment patches, apparently clusters of ocelli, discernible anterior to second pair of thoracic chaetal fascicles (Fig. 9E). Seven pairs of thoracic chaetal fascicles; thoracic membranes extend to of third pair of thoracic chaetal fascicles (Fig. 9E).
Many past publications expressed the belief that there is a single circumtropical species of Pomatostegus. However, according to ten Hove & Kupriyanova (2009: 78) there are three species. They consider Pomatostegus macrosoma Schmarda, 1861, a junior synonym of Terebella stellata Abildgaard, 1789, to be its type species. Ten Hove (pers. Apron absent. Last (6th) pair of thoracic uncinal tori located close together and meet, if not, almost meet mid-ventrally, and may have a very narrow flap posterior to them; when present, giving appearance of an apron (Fig. 9E).
Chaetae. Each bayonet-shaped chaeta (Fig. 10B) has a squarish basal boss and a long serrated blade, but lacks an unserrated notch. Thoracic uncini (Fig. 10C) possess 12 or 13 teeth in a single row and an anterior gouged process. Anterior abdominal uncinal tori are lateral (Fig.  9E) to somewhat dorsolateral, becoming more dorsolateral from about the 11th or 12th segment onwards. They are smaller than the thoracic uncini and bear 7 or 8 teeth in a single row (Fig.  10D). Posterior abdominal uncini (Fig. 10E) are still smaller but the number of teeth is somewhat higher, about 10.
Ontogenetic observations relating to operculum. One of the specimens provides evidence as to how opercular diameter increases in relation to increase in tube diameter as the worm grows in size (Fig. 10A). The earliest formed opercular tier is located at the distal end of the central column, and is the smallest in diameter. Each tier consists of a saucer-shaped part of light to darker coloured chitin-like material bear ing marginal scallops and a circle of darker spines just beneath it arising from the central column. The newly formed saucer-shaped tier is of the diameter of the now larger distal end of the opercular plate. Both, the newly formed saucer-shaped plate and the circlet of spines of the earlier-formed saucer-shaped part are enclosed within a transparent membrane. The latter extends from a rim-like extension beneath the tier (X) to the rim of the opercular plate (Y). As the worm grows, larger tiers are produced successively by the opercular disc to keep pace with the increasing tube diameter, while the central column also extends in length.
Remarks. Ten Hove (pers. com.) alerted the present author to the fact that the species frequently identified from the Indo-West Pacific as P. stellatus (Abildgaard, 1789) is actually P. actinoceros (Mörch, 1863). Reviewing published accounts P. stellatus, including that from the Indo-West Pacific, Pixell (1913: 78) states, under generic characters,: that the abdominal chaetae are "sickle-shaped (setae of Salmacina), and on p. 79: that she had "been able to find very few bladed sickles (setae of Apomatus) among the thoracic setae, the majority being rather short, still, simple bladed forms. The abdominal setae have the rounded teeth described by this author, but have nearly lost their sickle-like curve, this is, no doubt, due to their lengthened immersion in preservatives …".
Ten Hove & Kupriyanova (2009: 78) provide additional data with regard to opercular insertion, branchial eyes, and mouth palps. With regard to opercular insertion, however, they state that "from the fact that the first and second branchial radiole separated by the base of the peduncle, it is inferred that it is derived from the second normal radiole." Subfamily Serpulinae Macleay, 1840, sensu stricto (emended)
The subfamily diagnosis presented below takes into account the morphological characters of the extant genera constituting the Serpula-Spiraserpula-Crucigera-Hydroides clade (see Pillai & ten Hove, 1994).

Diagnosis.
Operculum usually present: inserted in position of second branchial radiole; always chitinous; peduncle nonpinnulate, may be pinnulate in early juveniles of certain taxa. Depending on taxa, consists of a funnel-shaped structure bearing raii ending in radial lobes only, rarely simple and with few or no radii and radial lobes; or a proximal funnelshaped structure with radii and radial lobes bearing a distal crown of spines (corona or verticil); or a funnel-shaped structure, bearing a number of blunt processes proximally below it. Inter-radiolar membranes absent; thoracic membranes present; apron, present in most taxa, absent in some as an adaptation to internal tube structures. Special collar chaetae frequently bayonet shaped, bearing a number of teeth on boss; may be reduced or modified in certain taxa. Thoracic uncini: saw-shaped in certain taxa, saw-shaped to rasp-shaped in others; most anterior uncinal tooth in thoracic and abdominal uncini larger than rest, simple. Distal ends of abdominal neurochaetae shaped like asymmetrically 'flattened trumpets', with one side longer than other, and bearing numerous serrations or denticulations. Tube not coiled; devoid of internal tube structures; or, coiled and bear internal tube structures.

Remarks.
Diagnoses of the individual genera are presented in turn on the basis of their type species as they are dealt with. The Genus Crucigera Benedict, 1887, type species Crucigera websteri Benedict, 1887, which is not present in the Kimberleys collection but occurs in Australian waters, is distinguished from the other members of the Serpula-Spiraserpula-Hydroides-Crucigera clade by its possession of characteristic processes or swellings proximally below its Serpula-like funnel-shaped operculum. Differences in form, number and arrangement of those proximal processes, besides other characters, are employed to distinguish between its species.

Hitherto unused morphological characters
Two hitherto unused useful additional morphological characters are employed in this paper, since they were found to be useful in distinguishing between apparently similar species.
The first relates to the unserrated notch occurring between the boss and the blade in bayonet-shaped special collar chaetae. Comparisons of the unserrated notch between several species of Serpulinae were undertaken by the present author (unpublished), including 17 in the present paper. It was found that the length of the unserrated notch relative to the serrated part of the blade or length of the teeth on the boss, is reasonably constant for each species, and can be useful in their identification. This is especially so in newly formed special collar chaetae with their bosses and blades partly or wholly still within the collar fascicle, since older chaetae could be abraded to various extents with use. Six character states can presently be recognized with regard that character. The first two relate to the absence of an unserrated notch, either due to: (a) the entire blade being serrated, e.g., in Serpula nudiradiata n.sp. (Fig. 30A-E), or (b) the entire blade being unserrated, e.g., in Spiraserpula caribensis Pillai & ten Hove (1994: figs 14K-M, 15B-I). The remaining four concern the length of the unserrated notch relative to either the length of the teeth on the boss, or the length of the blade, as follows: (c) short, i.e. shorter than or equal to the length of the teeth on the boss, e.g., in Spiraserpula snellii, Pillai & ten Hove (1994) and Fig. 35E-G in the present paper; (d) moderately long, i.e., longer than main teeth on boss, but up to about a third the length of blade as in Serpula (?) sp. (Fig. 31D-H), Spiraserpula zibrowii Pillai & ten Hove (1994: fig. 12F-I) and Spiraserpula plaiae Pillai & ten Hove (1994: fig. 13N-T); (e) very long, i.e. longer than a third of the length of the blade ( Fig. 17G-I) as in Hydroides pseudexaltatus n.sp.; (f) distally serrated, i.e., the entire blade beyond the boss, except for its distal end, is unserrated, a character state presently known to occur in only a single species, namely, Serpula narconensis Baird, 1864, (Fig. 22F in present account).
The second character relates to the gap separating the anterior abdominal tori of the two sides. In the present study it was observed that, during ontogeny, the anterior abdominal tori are initially short and located laterally in the earliest juveniles. As they grow older, additional uncini are added on at the medial end of each torus with the result that the tori gradually extend medially and become more dorsolateral in position. Depending on the species, as the worm grows older, the tori extend dorsally the corresponding pairs of the two sides may even almost meet or meet mid-dorsally. The uncinal tori of the more posterior abdominal segments however, do not increase in length as in the anterior segments, with the result that they gradually become more dorsolateral and lateral.
While abdominal uncinal tori occur laterally in the adults of certain species of Hydroides, more apomorphous conditions occur in others, as illustrated by the following examples. They are dorsolateral, separated by a wide gap in Hydroides kimberleyensis n.sp. (Fig. 14A); by a moderately wide gap in H. spiculitubus n.sp. (Fig. 20B); and by a very small gap in Hydroides adamaformis n.sp. (Fig. 11B). They almost meet along the mid-dorsal line in H. minax (Fig. 15B) and H. exaltatus (Fig. 12C).

Genus Hydroides Gunnerus, 1768
Type species. Hydroides norvegica Gunnerus, 1768. Diagnosis. Operculum chitinous; inserted in position of second radiole; consists of a Serpula-like proximal funnelshaped proximal part (or Latin equivalent infundibulum for brevity) usually bearing radii and radial lobes, surmounted by a distal crown (or Latin equivalent corona, or French verticil, for brevity) consisting of spine-like or other shaped processes. Number of thoracic chaetigers: seven. Special collar chaetae possess two to several teeth at base of blade and an unserrated notch of variable length relative to length of blade, depending on species. Most anterior uncinal tooth: larger than rest. Apron: present. Distal ends of abdominal neurochaetae shaped like asymmetrically "flattened trumpets", with one side longer than other, and bearing numerous serrations or denticulations. Tube: devoid of internal tube structures.
Remarks. The corona (or verticil) which consists of processes borne distally on the funnel-shaped operculum is an autapomorphous derivation of the condition in Serpula. Differences in its structure as well as of the marginal lobes of the proximal funnel-shaped part, besides other characters, are useful in distinguishing between species.
Tubes of many extant species Hydroides bear a pair of longitudinal ridges. The latter character could be useful in the determination of fossil Hydroides taxa, since the chitinous operculum is less likely to be found preserved.
With its numerous known species, and others still being discovered, Hydroides is currently one of, if not, the most successful among serpulid genera.

Description
Measurements. Maximum external width of tube, with flanges, 2.0 mm; total length of worm 20.0 mm, its thoracic width 1.0 mm; length of operculum and peduncle 4.7 mm, length of operculum alone 1.5 mm, length of radioles 2.7 mm, their pinnule-free tips about 0.6 mm, and length of abdomen 13.5 mm.
Tube. White, with two dorsolateral longitudinal ridges and a very wide flange on each side giving it a flattened trapezoidal cross-sectional appearance (Fig. 11A); longitudinal ridges separated by a wide, shallow, longitudinal groove between them; fine, incomplete, transverse grooves are present at irregular intervals between them; granular overlay absent. Worm. Radioles colourless; scattered brownish specks present between thoracic segments; a small patch of brown pigment present anteroventral to each thoracic uncinal torus; a small patch of brown pigment present at dorsal and ventral ends of each uncinal torus in abdomen.
Operculum in holotype on the right side, rudimentary operculum on the left; corona heterospinous; consists of 6 swollen processes (Figs 11A-E), each shaped roughly diamond shaped, arising proximally from its bottom angle, widening medially towards its lateral angles and curved towards its distal angle where it ends as ablunt point at the centre of the crown; its two lateral angles project as two short spines. Each process possesses a thick, transparent cuticle. One coronal processes is larger than the rest, but like the rest, curved towards centre of crown. All lack medial and basal processes. Infundibulum cup-shaped rather than funnel-shaped, its marginal lobes ending in T-shaped to crescent-shaped tips (Figs 11A-E). Small whitish specks scattered throughout infundibulum. Radioles: 11 on left, 10 on the right; their pinnule-free tips as long as and slender as pinnules (Fig. 11A Remarks. Hydroides adamaformis superficially resembles is H. tuberculatus Imajima, 1960(vide Imajima, 1960and Fiege & Sun, 1999: 121, fig. 11A-C). As described and figured in both latter publications, infundibular lobes of H. tuberculatus end in simple pointed tips. On the other hand, those of H. adamaformis possess T-shaped to crescentic tips ( Fig. 11A-E). The number of coronal processes in H. tuberculatus is 5; their shape is sub-triangular, and their cuticle is thick as described by Fiege & Sun (1999: 123). All the coronal processes in H. adamaformis, including the modified one, are diamond-shaped, and each bears a pointed spine on each lateral angle, while its distal angle is blunt and curved toward centre of the crown (Fig. 11A-E). On the other hand, all coronal spines in H. tuberculatus lack the pair of laterally directed spines occurring in H. adamaformis, whilst its modified coronal spine ends in a ventrally directed beak-like curved process (Fiege & Sun, 1999: fig. 11A). Although the number of coronal spines in H. adamaformis appears to be 5 from certain angles (Fig.  11B,C), it is actually 6 (Fig. 11D).
The external swelling described in Imajima's original account of H. tuberculatus (1976: 133, fig. 7a-c) is different: "each spine of the crown has a small external swelling at its shoulder, and a somewhat pointed accessory basal radial spine." As seen in Fiege & Sun's specimens of S. tuberculatus from the South China Sea (Fiege & Sun, 1999: fig. 11B) the distal ends of their sub-triangular unmodified coronal spines are quite stubby, somewhat inwardly distended, and bears an external anteriorly directed, transverse ridge-like swelling. The latter is the external swelling described by Imajima (1976) and as "a small chitinous knob on the external side of tip ( fig. 11A)" by Fiege & Sun (1999: 123). Hydroides adamaformis lacks such an external terminal chitinous knob on its coronal spines ( Fig. 11A-D).
Hydroides trilobulus  also has a remote resemblance to H. adamaformis. However, its corona consists of only three swollen processes, and their shoulders do not end in lateral spines, and its infundibular radii end in simple processes, whereas those of H. adamaformis end in swollen T-shaped to crescentic tips.

Measurements.
A maximum external diameter of 1.6 mm occurs in specimen from AM W202941. Longest available specimen has a total length of 16.8 mm and thoracic width of 1.0 mm; length of operculum and peduncle 4.2 mm, operculum alone, 2.0 mm; abdomen10.4 mm. Smallest specimen is a juvenile only 3.8 mm long with a thoracic width of 0.5 mm.
Tube. White, quadrilateral cross-sectional appearance and lacks a granular overlay (Fig. 12A,B,D); a pair of dorsolateral longitudinal ridges present, discontinuous in places; a thinner, more discontinuous, less developed median longitudinal ridge may or may not be present; transverse ridges present.

Worm.
A brown transverse pigment band present anterior to each thoracic uncinal torus. Coronal and infundibular spines end in transparent light-brown tips. A juvenile specimen in collection has two opercula ( Fig. 13L). Corona heterospinous; total number of coronal spines 6 to 10 in available specimens; in older specimens most dorsal spine enlarged, sub-terminally expanded laterally before ending in a ventrally curved hook (Figs 12E,F, 13A,B), narrower in younger specimens (Fig.  12C). A dimple may be present externally, on the subterminal enlargement (Fig. 12E,F). Remaining coronal spines curved outwards and tapered to a point. A short medial spine occurs halfway along each coronal spine, including the enlarged one (Figs 12F, 13A,B); may not be developed in juveniles. Infundibular radii end in simple pointed spinelike processes; number 22 in four juveniles. Infundibular radii end in simple pointed spine-like processes; number 22 in four out of the seven specimens, 23 in two, and only 14 in one of the two opercula of the juvenile described above.
Each infundibular spine appears jointed to the marginal lobe from which it arises, curved outwards, and tapers to a simple pointed tip in older specimens (Figs 12E,F, 13A,B); less obvious in juveniles. Radioles range from 7 per side in juvenile described above, to 14 on left and 13 on right in an older specimen; their pinnule-free tips very short, as slender as or thinner than pinnules. A pair of prostomial ocellar clusters present, purplish brown in colour, lighter in juveniles.
Chaetae. Special collar chaetae bayonet-shaped ( Fig.  13C-I), total number of teeth at base of blade 2-6; of which 2 may be large and the rest one to few smaller accessory teeth (up to 4 observed) at base of blade; two large worn out teeth seen in an older chaeta (Fig. 13C). A newly formed bayonet-shaped chaeta from within fascicle has 3 main teeth and a slightly smaller fourth one (Fig. 13E); another, 1 main tooth and 4 smaller teeth (Fig. 13 I). Unserrated notch in newly formed chaetae: moderately long, about one sixth of total length of blade (Fig. 13E); may be indistinct ( Fig.  13 I); blade long and tapered. Thoracic uncini saw-shaped: bear 7 teeth (Fig. 13J); anterior abdominal uncini similar, with 5 teeth in single row (Fig. 13K); posterior abdominal uncini rasp-shaped.
However, according to Imajima's description (1976: 127, fig. 4a-j), which does not include material from Eno-shima, each coronal spine, including the enlarged one, has a short medial spine close to its base. The Kimberleys specimens (Figs 12F, 13A,B) agree with Imajima's (1976) description. It appears, therefore, that there are two apparently similar species. The infundibular processes in both are simple, and the coronal spines are similar in that the enlarged pointed main spine is curved towards the centre of the crown, while the remaining ones are shorter, end in simple tapered tips, and are curved outwards.
Other collections identified as H. exaltatus were examined with a view to seeking an explanation for the above differences. A specimen, BMNH 1992.382, collected by Pillai in 1971 from Biak, West Irian, Indonesia, determined as H. exaltatus (Marenz) by ten Hove in 1988, is also similar to the older specimens from the Kimberleys with regard to this character. BMNH 1959.10.19.3-4, from Bougainville, Solomon Islands identified as H. exaltatus according to the label, contains 2 detached opercula and a detached corona. In the specimens from West Irian and Bougainville the enlarged coronal spine lacks the medial spine, while the unmodified coronal spines bear a short medial spine each.
Pillai's specimen from Sri Lanka, BMNH 1959.12.16.19, its identity confirmed by ten Hove in 1969, was re-examined during this study. Its enlarged coronal spine does not possess a medial spine, as in the above specimens from Bougainville. However, careful re-examination revealed that, of the remaining 7 coronal spines, the one next to the enlarged one lacks the short medial spine, another possesses only an insignificant stub, and each of the rest possesses a short medial spine. They are quite transparent and located close to the base of enlarged spine. Moreover, the adjacent medial spines meet and appear like demarcating lines between the bases of the main spines. It is evident, therefore, that the short medial spine on the coronal spines is a variable character in H. exaltatus. It may not be developed in some, or even in all the coronal spines; they may also vary in position, being closer to the base in some individuals than in others.
Another variation observed concerns bayonet collar chaetae. Although Imajima (1976) states that collar chaetae may bear one or two conical teeth at the base of the blade, they bear more than 2 in the Kimberleys material (up to 6 were counted), composed of 1 or 2 main teeth and the rest of smaller accessory teeth ( Fig. 13D-I).  (Marenzeller, 1885), from AM W21496. (A,B) two views of an operculum; (C-I) bayonet collar chaetae C, an older chaeta with worn out teeth. Remaining bayonet chaetae have 3 to 5 teeth at base of blade. Note that teeth do not extend way down the boss as in H. pseudexaltatus (Fig. 17E,F,G-I), and their un-serrated notch is moderately long compared to that in H. pseudexaltatus in which it is very long.]; (J) thoracic uncini; (K) anterior abdominal uncini; (L) a juvenile specimen; it has two opercula; the main spine in the better-developed operculum is directed inwards, while it is directed outwards in the other; its anterior abdominal uncinal tori are lateral, separated by a wide gap dorsally, compared to older specimens (Fig. 12C) in which they almost meet mid-dorsally.

Description
Measurements. Apparently a very small species; maximum external tube diameter: 1.2 mm in holotype, 1.0 mm in paratype. Total length of longer of the two worms is 10.0 mm, the other 9.1 mm. However, latter sexually mature, since ova being present in posterior part of abdomen. Thoracic width: only 0.5 mm in both. Length of operculum and peduncle 2.5 mm in first specimen, 1.7 mm in the second; length of abdomen, 5.1 mm long in first specimen, 7.3 mm in second. Number of abdominal segments: 66 in first, 71 in second; capillaries starting from 63rd segment in both.
Tube. Light caramel coloured in holotype, white in paratype; lined internally by chitin-like material, light brown in the former, more conspicuous and darker brown in the latter. It has a characteristically wide, apparently flattened dorsal area with a longitudinal ridge on each side, sloping sides, and a flange on each side, giving it a trapezoidal cross-sectional appearance (Fig. 14Q). Fine transverse ridges traverse the gap between the two longitudinal ridges, most of them not extending beyond the wide mid-dorsal longitudinal groove itself, thus leaving the longitudinal ridges comparatively clear. Irregular transverse ridges present along the flanks; a granular overlay absent.
Worm. Posterior part of collar and ventral side of thorax brown; a transverse brown band present anterior to each thoracic uncinal torus; brownish pigmentation present transversely across ventral side of abdominal segments, fainter bands dorsally in posterior part of abdomen, latter more pronounced in second specimen. Infundibular lobes are brown distally, their tips blackish; opercular peduncle whitish.
Operculum on the right side in both specimens; number of radioles radioles 8 per side, their pinnule-free tips as long as and as thick as pinnules (Fig. 14A,L,Q). Corona of both specimens has 5 triangular, brownish, chitinous processes each, which arise directly from the infundibulum, i.e., not collectively borne on neck-like part arising from it ( Fig.  14A-D,L-P,Q); their blunt tips directed outwards. Diameter of coronal base somewhat greater than or equal to that of infundibulum, so that, when viewed from anterior end, only the pointed tips of the marginal infundibular processes may be seen, if at all (Fig. 14D,N). Coronal processes ( Fig.  14A-D,L-P,Q) translucent light brown, their tips and edges transparent light brown to honey coloured; each roughly triangular in cross-section, thick and ending distally in a bluntly pointed transparent tip; also possesses, medially, at it base, 4-5 tiny processes (Fig. 14D). Coronal processes are all similar in holotype (Fig. 14B,C); in the paratype, the translucent part at the tip of one of the coronal spines bears a short, sharp, translucent, medial sub-terminal spine (Fig. 14 O,P), which appears to represent the fully developed condition.
Infundibular lobes minute, acutely triangular, and end in outwardly curved pointed tips ( Fig. 14A-C,L-P,Q). They number 24 in the holotype, difficult to count in the paratype. Two faintly brown clusters, of what appear to be prostomial ocelli, present in the first specimen, darker and clearer in the second. Thoracic glands: not discernible. Anal segment appears to bear a pair of ocellar clusters; they are clearer in the second.

Chaetae.
Each collar fascicle bears about 5 bayonet-shaped special collar chaetae. Two newly formed ones from within the fascicle possess a small accessory tooth between two larger conical teeth (Fig. 14E,F); remaining three older, their blades and accessory tooth and conical teeth at the base of blade worn out ( Fig. 14G-I). In newly formed bayonet chaetae, the unserrated notch is comparatively short, about one-tenth to about one-twelfth of the total length of the blade (Fig. 14E,F). Thoracic uncini (Fig. 14J) and anterior abdominal uncini (Fig. 14K): saw-shaped, bear 6 and 5 teeth, respectively; posterior abdominal uncini rasp-shaped. Abdominal uncinal tori: 15 or so pairs of anterior abdominal tori are dorsolateral, with a wide gap separating the tori of the two sides ( Fig. 14A).

Remarks.
Hydroides kimberleyensis is somewhat similar with regard to its small size and characters of its coronal processes to H. xishaensis , H. trilobulus Chen & Wu, 1978, and H. rhombobulus Chen & Wu, 1980, but there are important differences among them. Total lengths of their holotypes are, only 4.8 mm for H. xishaensis and 4.0 mm for H. trilobulus, while it is 9.0 mm for H. rhombobulus Chen & Wu, (1980: 250). Total lengths of the two specimens of H. kimberleyensis are 10.0 mm and 9.1 mm, respectively. In spite of their comparatively minute sizes, the holotype of H. xishaensis has a total of 20-24 branchial radioles (10-12 per side), while that of H. trilobulus has a total of 20-26 (i.e., 10-13 per side). On the other hand, the holotype of H. kimberleyensis has only 16 radioles (8 per side). The 9.0 mm holotype of H. rhombobulus has 12 radioles (6 per side), somewhat similar to H. kimberleyensis in this respect.
The number of coronal processes in both specimens of H. kimberleyensis is 5, but it is 4 in H. rhombobulus. Moreover, the colour and form of coronal processes in H. kimberleyensis and H. rhombobulus are different. In H. kimberleyensis they are brownish, triangular, convex on the outside, with transparent edges and tips, and each bears 4-5 tiny basal processes (Fig. 14D). In H. rhombobulus, on the hand, the coronal spines are rhomboidal in shape and apparently lack medial basal processes. In H. kimberleyensis, the diameter of the infundibulum and infundibular processes is nearly equal to that of the corona, the infundibular lobes slender, and their pointed tips pressed closely against the coronal base ( Fig.  14A-D,L-Q,). In H. rhombobulus, on the other hand, the diameter of the infundibulum and infundibular processes is greater than that of the corona, and its infundibular lobes are directed outwards (Chen & Wu, 1980: pl. 1, fig. 1). Minor differences also appear to exist in their chaetae. Bayonet shaped collar chaetae of H. kimberleyensis typically possess two conical processes at the base of the blade, with a smaller accessory tooth between them, but in H. rhombobulus they have been described and figured as possessing only two conical processes. Thoracic and abdominal uncini of H. kimberleyensis bear 6 and 5 teeth, respectively, whereas they bear 7 and 6, respectively, in H. rhombobulus. Hydroides xishaensis  (plate 1, fig. 1) has 5 triangular, concave coronal processes, lacking basal spines, and their pointed tips are directed towards the centre of the corona like claws of a grapple-iron. As also shown in Chen & Wu's figure the infundibular processes and their triangular tips are directed outwards. In contrast, the coronal processes of H. kimberleyensis are not clawed and directed inwards but outwards at their tips, and each one of them bears a number of basal spines (Fig. 45D). Moreover, its infundibular processes are pressed against the bases of its coronal spines, and only their tips are curved outwards. An additional character to support the distinction between the two species concerns differences in the unserrated notch in the bayonet-shaped special collar chaetae. As figured by Chen & Mu (1978: plate 1, figs 2,3), it is absent or short. On the other hand, it is moderately long (longer than the teeth on the boss, but up to one about 1 ⁄3 the total length of the blade) in H. kimberleyensis.
Unlike in many species of Hydroides dealt with in this paper, anterior abdominal tori are lateral, with a wide middorsal gap separating them.

Hydroides minax (Grube, 1878)
Worm. Operculum blackish in larger specimen AM W21445, translucent white with faintly yellowish spines in the second. Number of coronal processes 7 in former ( Fig. 15D), and 6 and 7 in the two opercula of second, AM W202951 (Fig.  16C); all devoid of basal processes. Most dorsal coronal process stout, elongated, and terminates in three strong sharply pointed hooks, one curved ventromedially and the other two curved posterolaterally (Figs 15B-E, 16C); rest simple, outwardly curved spines. First specimen has the higher number of infundibular lobes, 24. They are simple and end distally in a swelling (Figs 15B-E, 16C) which may be somewhat T-shaped (Figs 15B, 16C). Maximum number of radioles per side in the two specimens is 21; their pinnule-free tips short, slender; only slightly thicker than the pinnules. A blackish cluster prostomial ocelli present on each side, anterior to the second thoracic chaetiger (Fig. 15B,C). Abdomen deeply folded along mid-ventral line and its cuticle appears to be transversely sloughed off as spicules (Fig. 15C).
Chaetae. Special collar chaetae bayonet-shaped (Figs 15F,G, 16D-F), bear 2 conical processes at base of blade. Thoracic uncini saw-shaped, with 7 teeth in single row ( Remarks. Pixell (1913), Fauvel (1953) and Straughan (1967a,b) distinguish between H. minax (Grube, 1878) and H. monoceros , on the basis that the corona of the latter arises asymmetrically from the infundibulum in the former but symmetrically in the latter. Pillai (1971) mentioned the difficulty of distinguishing between them, and the need for further studies to specify the differences. Hartman (1959) believed that they may be synonymous. Imajima (1976Imajima ( , 1982 and Imajima & ten Hove (1984, 1986) considered them synonymous, as also supported in the present study. Variations in their coronal spines range from those occurring in juvenile specimens of H. minax to those in older specimens of H. monoceros. They are translucent, somewhat greenish-yellow in colour, and the most dorsal coronal spine is very short in juveniles, whereas in older specimens they are brownish and the most dorsal spine is comparatively very long. The specimen with the longest modified coronal spine examined in the study is in Crossland's collection from Zanzibar. Some of its measurements are as follows: maximum tube diameter 3.1 mm; length of operculum and peduncle 6.0 mm; length of operculum including long stiff constriction below the infundibulum and the peduncle 4.2 mm, and length of most dorsal spine 3.0 mm. The tips of infundibular spines may be more or less swollen, irrespective of the age and size of the worm. Sloughing off spicules of ventrally along the abdomen does not appear to occur to the same extent as in Hydroides spiculitubus n.sp., described below.
Hydroides pseudexaltatus n.sp.  Tube. White, maximum external diameter at its anterior end 2.7 mm; its surface is irregular, but two somewhat rounded and irregular longitudinal ridges with a moderately wide shallow longitudinal gap between them can be recognized, and an irregular flange on either side ( Fig. 17A) which gives it a roughly trapezoidal cross-sectional appearance. Irregular transverse ridges present, also between the longitudinal ridges. Fractured end of tube reveals a moderately thick caramel to light brownish translucent inner layer, with patchy brownish pigment inside.
Worm. Coronal and infundibular processes transparent light yellow and with brownish tips. Corona heterospinous and consists of 9 spines, all devoid of medial, lateral and basal processes (Fig. 17D); enlarged dorsal spine somewhat swollen in middle and ends distally in a hook curved towards Etymology. Named after the superficial similarity of its operculum to that of H. exaltatus (Marenzeller, 1884).

Description
Measurements. Worm incomplete posteriorly; width of thorax 0.5 mm; length of operculum and peduncle 2.5 mm, of operculum alone 1.0 mm.

Tube. Missing.
Worm. Thorax has dark brown pigmentation both dorsally and ventrally, and anterior to the thoracic tori; modified coronal spine translucent and off white, remaining coronal spines translucent and light brown. Modified dorsal spine vesicular, shaped like a swollen beak, curved towards the centre of the crown (Fig. 18A,B-F); remaining coronal spines number 6, highly reduced, pressed against base of the modified spine; their bases broad, but their short, simple, and Remarks. Hydroides simplidentatus n.sp. is somewhat similar to H. brachyacanthus Rioja, 1941. Sample BMNH 1933.445 collected by C. Crossland from Coiba Island, Pacific coast of Panama, and determined as H. brachyacanthus Rioja, 1941, by H. Zibrowius in 1970, was compared with the present species. Although its corona is superficially somewhat similar to that of H. simplidentatus ( Fig. 38K-P), in that a modified spine is enlarged swollen and curved towards the centre of the corona, close examination shows that they are quite different. In H. simplidentatus it is vesicular, like a parrot's beak, hooked towards the centre of the crown, and enclosed in a translucent covering ( Its modified coronal process lacks one, and there is no room left for them at the bases of the extremely reduced unmodified coronal processes which occur tightly pressed to the base of the modified coronal process. Straughan (1967a) reports S. brachyacanthus from east ern Australia, also incorporating material described by Dew (1959) under the same species. The latter also superficially resembles H. brachyacanthus and H. simplidentatus, but it is different from both of them. A collection by Dew from Townsville, Queensland, BMNH 1959.10.19.61-62, and determined as H. brachyacanthus, sensu Dew, by H. Zibrowius in 1970, was examined during this study. In one specimen, 4 of the more dorsal coronal spines are larger that the rest and curved over the centre of the crown. One of the latter, in turn, is only slightly larger than the rest. However, Dew (1959) states that the crown consists of nine "stout inwardly-pointing spines, largest and stoutest of which curve over remaining eight, forming a cage." Dew's (1959) fig. 7B shows a single enlarged spine. It is evident that one or more enlarged coronal spines may be curved over the centre of the crown. While a shoulder can be faintly recognized in the four larger spines, they are strongly developed in the remaining coronal spines, being tooth-like as described by Dew (1959) and Straughan (1967a). Firstly, its main process is a strong curved spine (vide Dew, 1959: fig. 7B). Secondly, as described by Dew, "remaining spines, except pair flanking large spine, have a blunt, tooth-like projection on shoulder of curve". Straughan's (1967a) description is similar. In contrast, both H. brachyacanthus Rioja, 1941 and H. simplidentatus a single enlarged process curves over the centre of the crown. Further studies are required to determine whether Straughan's Eastern Australian material belongs to a species different from H. brachyacanthus Rioja.
Etymology. The specific name simplidentatus refers to the simple unmodified spines at the base of the enlarged coronal spine. other Hydroides species, is the presence, in all three specimens, of a detachable inner tube (Fig. 19D). Under high examination, consists of transversely arranged, closely adposed, spicules (Fig. 19D,E) sloughed off from the ventral abdominal wall (Fig. 20A).

Worm.
A brown pigment patch may be present in constriction between operculum and peduncle (Fig. 20C); ventral side of thorax and edges of thoracic membrane brown; a brown pigment band present anterior to each thoracic uncinal torus (Fig. 20A). Abdomen: a narrow pigment band present anterior to each torus, ending ventrally in a brown patch; a similar pigment band present across ventral side of each segment, joining corresponding ventral patches of both sides. One or two opercula may be present. Tips of coronal spines and their lateral spines, black. One specimen has a single operculum (Fig. 20B), the other, two well-developed opercula each (Fig. 20A). Three out of the 5 opercula among them, bear 7 coronal spines, the remaining two bear 6. All spines are similar (homospinous) and tapering; initially directed anteriorly and inclined towards the centre of the crown rather than outwards, then curved outwards before ending in curved pointed tips (Figs 20A-C, 21A-E). Located a little less than halfway along each spine and arising from it, externally, is a pair of short outwardly directed spines (Figs 20B,C, 21A-E). A medial spine is, present arising a little distal to the pair of lateral spines, as well as a basal process, both sharply hooked towards the base of the crown (Fig. 21D,E). Infundibulum: out of the five opercula amongst the three specimens one has 31 marginal processes, two 28, one 27, and one 26. Infundibular spines: transparent, their tips light brown; each marginal process has a sub-terminal knee-like bend towards centre of crown, beyond which it tapers to a simple pointed outwardly curved process (Fig. 21C). An aberrant condition occurs in one specimen, in which a single marginal infundibular process in one of its opercula is forked. Maximum number of branchial radioles counted per side: 9 in all three specimens. However, specimen separated from AM W21390, has 11 radioles per side; their pinnule-free tips slender, and increase in length from ventral to dorsal radioles; about one-sixth of total length of radiole.
A light-brown cluster of prostomial ocelli present on each side, just anterior to the first pair of uncinal tori. Thoracic glands: absent. Anal segment bears a pair of brown to blackish clusters of ocelli ( Fig. 20A,B,D,G).

Remarks.
The new species H. spiculitubus from the Kimberleys is only superficially similar to H. tambalagamensis from Sri Lanka with regard to the operculum. This has apparently led to the latter being reported from Southwest Japan by Imajima (1976: fig. 2a), and by Chen & Wu (2000: fig. 6K,L) from the waters off China.
Both species are similar in that their coronal spines are homospinous, and each bears a pair of short pointed spines located about halfway along its length, a short, inwardly curved, medial spine a little beyond them, and a basal spine. However, differences between them, following reexamination of the type specimen of H. tambalagamensis, BMNH 1960.3.13.30, are as follows. The coronal spines of H. spiculitubus are comparatively slender, directed anteriorly and meet sub-terminally, before curving outwards. In comparison, those of H. tambalagamensis are stout at their bases, rapidly tapering, directed outwards, and do not meet sub-terminally (vide   fig. 12A,B). The pair of spines in H. spiculitubus are external ( Fig. 21A-C), whereas they are lateral in H. tambalagamensis, one on either side of each coronal spine   fig. 12A,B). The medial spine of each coronal spine in H. spiculitubus is sharply hooked towards the centre of the coronal spines (Fig. 21D,E), in Imajima's Southwest Japan specimens (Imajima, 1976: fig. 2b) and in Chen & Wu's specimens form China (2000: fig.  6K). On the other hand, it is directed almost at right angles to each main spine in H. tambalagamensis is  fig . 12A,B). The basal process of each coronal spine in H. spiculitubus is slender, and acutely hooked towards the centre of the corona (Fig. 21D,E), whereas it is is obtusely triangular and blunt in H. tambalagamensis   fig. 12B).
A further, very important difference between the two species is with regard to pinnule-free tips of radioles. In H. spiculitubus they are soft and similar in texture to that of pinnules (Fig. 20A,B), as also in almost all other serpuli morphs. However, as confirmed after re-examination of the type specimen during this study, each radiole in H. tambalagamensis, terminates in a stiff, chitinous colourless, inwardly curved, pointed hook (vide Pillai, 1961: 38, fig. 12C).
Anal eyespots are present in H. spiculitubus, but were not observed in the Southwest Japan material by Imajima (1976), material from China by Chen & Wu (2000), and H. tambalagamensis by .
As described earlier, there are two important differences between the two species with regard to the tube. The first is the characteristic detachable inner tube of H. spiculitubus, composed of more or less transversely adposed chitinous spicules sloughed off from the ventral side of its abdomen (Fig. 19D), the function of which is obscure. In contrast, H. tambalagamensis has no such inner tube consisting of chitinous spicules as originally described by . Imajima (1976) and Chen & Wu (2000) do not mention it in their descriptions of "H. tambalagamensis". The second difference concerns the pair of longitudinal ridges along their tubes. In the material described as "H. tambalagamensis" from Southwest Japan by Imajima (1976: 123, fig. 2a-j) they are dorsolateral, separated by a relatively wide and shallow gap, which may also vary somewhat in width in different parts of the tube. Its granular overlay is concentrated along the longitudinal ridges and lateral edges of the tube. In H. tambalagamensis , on the other  hand, and as confirmed after re-examination of the holotype during the present study, its two longitudinal ridges run along the crest of the tube rather than dorsolaterally, they are comparatively thicker, the longitudinal groove between them is narrower and deeper, as figured by   fig. 12G). Its entire tube surface has a conspicuous granular overlay.
Etymology. Named after its characteristic inner tube constructed from chitinous spicules sloughed off from ventral side of the abdominal wall, occurring within the normal calcareous tube, within which the worm lives (Fig. 19D,E).

Description
Measurements. Holotype: maximum external diameter of tube 1.2 mm; total length of worm 10.9 mm; thoracic width of 0.6 mm; length of operculum and peduncle 2.8 mm, operculum alone 1.0 mm; abdomen 6.5 mm, total number of segments 73, capillaries start from the 57th segment. Paratype: worm only only slightly smaller, 10.3 mm long, its thoracic width 0.4 mm, length of operculum and peduncle 2.7 mm, length of operculum alone 1.0 mm, length of abdomen 5.8 mm, and number of abdominal segments about 58; capillaries start from about 43rd segment.
Tube. Spirally coiled, in both holotype and paratype (Figs 22A, 23A), unlike in other hitherto known species of Hydroides. Coils laid obliquely upon each other in holotype (Fig. 22A), upon each other in paratype (Fig.  23A); bonded together in both; small tubeless area or umbilicus present in coil centre of both, similar to that occurring in certain species of Spiraserpula Regenhardt, 1961(vide Pillai & Hove, 1994. In holotype, outer part of each coil white, its inner part consisting of a semitranslucent caramel-coloured band (stippled areas in Fig. 22A); paratype similar (stippled areas in Fig. 23A), but colour of its inner band lighter, almost white. Two longitudinal ridges may be discernible in certain parts of the tube where it may give it a more or less squarish cross-sectional appearance. A granular overlay present, more prominent along the longitudinal ridges.
Worm. Corona consists of 9 processes in both specimens; its modified dorsal process small, short, vesicular, and with a dorsolateral extension on each side and a medial swelling; T-shaped when viewed from anterior end (Fig. 22C,E); in lateral view, medial swelling is anteroventrally directed (Fig.  22B,D). Remaining coronal spines, as well as infundibular processes, end distally in somewhat T-shaped swellings (Fig. 22B,D,E). Holotype has higher number of infundibular processes, 25, the paratype 21. Radioles number 8 and 7 in holotype, 7 pairs in paratype, their pinnule-free tips long and conspicuous, about one-fifth the length of radiole ( abdominal neurochaetae almost capillaries, except that their tips are serrated along one side (Fig. 23J) as in many Serpulinae s.str.

Remarks.
Like Hydroides spiratubus n.sp., three other species of Hydroides are known to possess a modified coronal spine that is T-shaped in crosssectional appearance, together with unmodified coronal spines ending in swollen to T-shaped tips. They are, H. albiceps (Grube, 1870), H exaltatus var. vesiculosus Fauvel, 1923, from Îles Gambier, French Polynesia, andH. trivesiculosus Straughan, 1967, from Queensland. Ten Hove kindly sent his extensive original notes and drawings on several collections of the above species, including those of his reexamination of specimens and type specimens described by various authors from Elat in Israel to locations in the Indo-West Pacific. They will be briefly referred to in the following discussion. Willey (1905) and Pillai (1960) describe H. albiceps from Sri Lanka. However, the latter appears to be different from those those studied by ten Hove. Willey describes the coronal processes of a small specimen, only 7.0 mm long, from Galle (1905: 312, pl. VII, figs 180, 180a) as follows. It consists of "eight nearly erect slightly curved virgulae and a laterally compressed ovate lamina dorsalis, the latter being a direct continuation of the columella and bearing a pair of broad dorsolateral chitinous hamuli." Willey uses the diminutive plural "hamuli" of the Latin word "hamus", meaning hook, to describe the small lateral spine on each shoulder of the modified coronal process. He describes the corona of a larger specimen, 20.0 mm long as consisting of 7 processes, and that the modified coronal process (lamina dorsalis) as being "quite flattened except at the back." According to Willey's figure (1905: pl. VII, figs 180, 180a) the unmodified coronal processes are outwardly curved and pointed at their tips, and its infundibular processes do not end in swollen to T-shaped tips. Its modified coronal process is laterally flattened, except dorsally where it bears a pointed laterally directed "chitinous" spine on each shoulder. Pillai's description and figures (1960: fig. 4F-M), based on a collection of 33 specimens from the Pearl Banks, agrees with that of Willey (1905) with regard modified and unmodified coronal processes and the infundibular processes. Tubes are "quadrilateral in cross-section", the attached side being wider than the flattened opposite side, and the two upper angles of the latter being sometimes drawn out into low lateral longitudinall ridges. No spirally coiled tubes were observed. Bayonet-shaped special collar chaetae possess two conical processes at the base of the blade. Throracic uncini bear about 7 teeth in a single row.
Although the collections studied by ten Hove: PHYT 3436 from El Bilyayim, Gulf of Suez, and the following collections from Elat: NS 2149; 61-73; NS2148 RS TAV; NS2142 TAV; NS2145 TAV, and Elat8 are superficially similar to those described by Willey and Pillai from Sri Lanka, there are three differences between them. Firstly, the modified coronal process of the former does not possess the pair of pointed mediolateral spines that occur in the latter. Secondly, the unmodified coronal spines of the former end in swollen to T-shaped tips, whereas they are simple, outwardly directed and pointed in the latter. Thirdly, the tips of the infundibular radii in in the former vary from being pointed to somewhat swollen, whereas they are simple and pointed in the latter. The numerous specimens of H. albiceps examined and figured by ten Hove from collections in the AM, Sydney, from Australia (Lizard Island, Townsville, Barney Point, Gladstone, Shoal Point, Tannum Sands, Burleigh, Cairns, Pats Point, Mackay, Noosa Heads, York Peninsula and Port Moresby), and Ambon, also agree with the non-Sri Lankan collections discussed above. Ten Hove does not describe spirally coiled tubes in any of the collections examined, a characteristic of H. spiratubus n.sp. But according to his drawings of the tubes from AM W3975 (Barney Point, Gladstone) AM W3960 (Shoal Point) and AM W3961 (Tannum Sands) are trapezoidal in cross-section and bear a pair of longitudinal ridges.
The second species that needs to be compared with H. spiratubus n.sp., is H. exaltatus var. vesiculosus Fauvel   1923: 50, 1953: 641) from French Polynesia. Fauvel distinguished it from H. exaltatus (Marenzeller, 1884) on the basis that its modified coronal process is "a large hollow vesicle" which "takes the place of the great unpaired hook." Fauvel (1923) adds that H. exaltatus var. vesiculosus "is a connecting link between H. exaltatus (Marenzeller) and H. albiceps (Ehrenberg)." However, Fauvel's specimens from French Polynesia differ from both H. spiratubus and H. trivesiculosus in having a significantly higher numbers of coronal processes. According to Fauvel (1923: 49), they number about a dozen "Les dents du calice supérieur, au nombre d'une douzaine, sont subégales recourbées en dehors et terminées croc émoussé sans pointes laterales." The third species that needs to be compared with is H. trivesiculosus Straughan, 1967. The corona of the former (Fig. 22B-E) consists of 9 processes (in both holotype and paratype). Its modified dorsal process is small, short, vesicular, and has a dorsolateral extension on each side as well as a medial swelling. Its unmodified coronal spines end distally in swollen tips. On the other hand, based on Straughan's study of 6 specimens (1967: 34, fig. 3H,I), the corona of H. trivesiculosus, consists of only 5 processes. The modified process is massive, trivesicular in cross-sectional appearance and elongated, while the 4 unmodified processes are like tiny appendages at the base of the latter, ending in minute pinhead-like swollen tips.
Ten Hove's studies and drawings referred to above include collections both identified and misidentified as H. tivesiculosus Straughan, 1967, as fig. 7) describes that of the French Polynesian species as being loosely coiled: "tube blanchâtre, faiblement bosselé, sans ornementation speciale, est lâchement enroulé sur la coquille d'une Huître perlière." Tubes of many serpulids may be sinuous or loosely coiled. Tube of "H. albiceps" from Sri Lanka, as well of the above collections studied by ten Hove, is quadrilateral to trapezoidal in cross-section and usually bears a pair of longitudinal ridges.
Etymology. Named after the spirally coiled tube.

Description
Measurements. From AM W21412: largest tube diameter 1.5 mm; thoracic width in three specimens 1.0 mm; maximum; length of operculum and peduncle 3.7 mm; length of operculum only 0.9 mm. Maximum number of branchial radioles observed 10 per side. Number of abdominal segments could not be determined since worms are incomplete posteriorly; however, 25 could be counted in the available anterior part in one specimen. Total length of highly coiled worm from AM W202943 without its tube: 9.9 mm; width of thorax 1.0 mm, length of operculum and peduncle 3.0 mm, length of operculum 1.3 mm, length of abdomen 5.3 mm, and number of abdominal segments about 55. A smaller specimen 4.8 mm long has 51 abdominal segments, with capillaries commencing on about the 40th segment.
Tube. White in juveniles, dull white in older specimens; squarish to trapezoidal in cross-section; possesses two longitudinal ridges with a comparatively wide longitudinal depression between them (Figs 24A,B, 26A-C), which may only be faintly developed (Fig. 24E). A granular overlay is present along certain areas (Fig. 24A,B,E); may be faint in some specimens.
Worm. Operculum on either left or right side; light caramel coloured, its cuticle transparent. Number of coronal spines range from 11-15; number of infundibular lobes 17 to 24. Coronal and infundibular spines show variations associated with ontogenesis. In infundibular spines of early juveniles (Fig. 24D,F), their tips are simple and pointed. Modified modified coronal spine becomes larger than the rest, develops a translucent medial swelling, and ends in a pair of blunt outwardly directed processes (Fig. 24C,D). Unmodified coronal spines: outwardly directed and devoid of basal spines. Infundibular spines may remain simple and pointed while coronal spines become swollen at their tips, foreshadowing the T-shaped tips of later stages, while the medial swelling and forked tips of the modified coronal spine become more prominent (Fig. 24F). In a later stage (Fig.  24G), both infundibular and coronal spines are T-shaped; and medial and the two terminal processes of the modified coronal spine become even more prominent (Fig. 24G).
The adult worm shown in specimen from AM W202943 (Fig. 25A) has 9 radioles plus rudimentary operculum on left, and 8 plus operculum on right. The smaller specimen mentioned under measurements has 6 radioles per side. Although 11-15 coronal spines were counted in juveniles, only 10 are present in the present adult specimen, making their range in the present collection from 10-15; all devoid of a basal process. The medial process of modified coronal spine is a firm, beak-like process, curved towards the centre of the crown (Fig. 25A-C). Both processes at the distal end of the modified coronal spine are fully developed, their tips pointed and laterally curved ( Fig. 25A-C). Infundibular radii number 19. While the tips of the unmodified coronal spines and the marginal infundibular processes are T-shaped or end in rounded tips in later juveniles, they are anchor-shaped and with pointed transparent laterally directed tips in adults ( Fig. 25A-C).
Chaetae. Bayonet-shaped special collar chaetae (26G-J) possess two main conical processes at the base of the blade and 1 or 2 small accessory teeth between them; unserrated notch short, about ¼ length of blade. Thoracic uncini: sawshaped, with 6-7 teeth in a single row (Fig. 26K); anterior abdominal uncini similar, with 4 or 5 teeth (Fig. 26L); posterior abdominal uncini rasp-shaped. Anterior abdominal uncinal tori of the two sides are separated by a narrow gap (Fig. 24C); which is wider in juveniles (Fig. 26D).
Remarks. The new species, Hydroides trihamulatus is apparently similar to H. malleolaspina Straughan (1967: 222, fig. 7a-f), originally described from Queensland, with regard to the T-shaped infundibular and modified coronal spines. However, it is different with regard to the modified coronal spine. Straughan (1967) describes the latter in H. malleolaspina as being hammer-shaped distally and lacking a medial process, whereas, in H. trihamulatus, it ends in a bifid tip and bears a medial hooked process. Although the medial process and both processes at the distal end of the modified coronal spine are soft and blunt in juveniles of the present species (Fig. 24C-F), they are firm and beak-like in adults ( Fig. 25A-C). Hydroides rectus Straughan (1967: 225, fig. 8a-f) differs from both H. malleolaspina and H. trihamulatus in that its unmodified coronal spines and infundibular spines are not T-shaped but tapered and pointed distally, as described and figured by Straughan (1967: 225, fig. 8b-d). Moreover, the modified coronal spine of H. rectus is not bifid at its tip as in H. trihamulatus, but has a process at its distal end, which appears to be at right angles to it and directed towards the centre of the crown.
Etymology. The specific name trihamulatus refers to the 3 chitinous hooks borne by the modified coronal process in adults, one medial and curved towards the centre of the crown, and the other two located terminally and directed laterally.

Description
Measurements. Maximum external diameter of tube 3.7 mm, width of thorax 1.2 mm, length of operculum and peduncle 5.7 mm, length of operculum 2.1 mm, diameter of operculum 1.5 mm, width of thorax 1.2 mm; worm incomplete posteriorly.
Tube. White; coiled upon itself; possibly with a non-coiled part preceding the coiled part; mutually bonded coils; a conspicuous translucent granular overlay present (Fig. 27A) and two rounded longitudinal ridges.
Worm. Holotype, operculum on right, rudimentary operculum on left (Fig. 27B); operculum bell-shaped, its proximal part rounded; number of radial lobes comparatively small, 17; broad, rounded, and with a thick translucent cuticle (Fig. 27B,C); tubercles absent; constriction between operculum and peduncle sharp. Number of radioles: 17 on each side, their pinnule-free tips not significantly different from pinnules (Fig. 27D). A large cluster of reddish prostomial ocelli present on each side. Abdomen: incomplete posteriorly.
Chaetae. Bayonet-shaped special collar chaetae (Figs 27E-I, 28A,B) bear two large acutely triangular to conical teeth at the base of blade; some bayonet-chaetae may have a smaller accessory tooth in between (Fig. 27F,G). Unserrated notch: short in older bayonet chaetae (Fig. 27E-I), moderately long in newly formed ones within collar fascicles (Fig. 28A). Thoracic and anterior abdominal uncini: saw-shaped, bear 4 or 5 teeth (Figs 27J,K, 28D,E). Anterior abdominal uncinal tori: dorsolateral, with a wide gap separating corresponding pairs; gap about half the length of an uncinal torus in the region of the 6th to 10th abdominal segments (Fig. 28C).
Remarks. The differences among Serpula amplilobata and the other species of Serpula dealt with in this paper, namely, S. jukesii Baird, 1865, S. nudiradiata n.sp., and S. watsoni Willey, 1905, are as follows. (a) Its tube is coiled upon itself and mutually bonded together, whereas those of S. jukesii, S. nudiradiata, and S. watsoni are not. (b), The operculum in S. amplilobata and S. watsoni is bell-shaped, whereas that of S. jukesii is funnel-shaped. (c) The operculum of S. amplilobata bears a few comparatively large radial lobes, whereas that of S. nudiradiata, S. jukesii and S. watsoni bears a relatively large number. (d) The opercular cuticle in S. amplilobata is thick and translucent, as seen in the grooves between and tips of opercular radii, whereas it is not so in S. nudiradiata and S. jukesii and S. watsoni. (e) The opercular radii of S. amplilobata and S. nudiradiata are devoid of tubercles, whereas those of S. jukesii and S. watsoni bear tubercles. (f) The proximal part of the operculum in S. amplilobata and S. watsoni is bulbous, whereas that of S. jukesii is tapering (as in other Serpula spp. with funnel-shaped opercula). (g) The constriction between the operculum and peduncle is sharp, in S. amplilobata and S. watsoni, whereas whereas it is not so in S jukesii.
Etymology. The specific name amplilobata is derived from the Latin amplus = large and the Greek lobos = lobe, to describe the large rounded marginal lobes at the end of the opercular radii. They are few in number and much larger in size compared to those in species such as S. jukesii and S. nudiradiata.

Description
Measurements. Tube of specimen in AM W21934 has the larger maximum external diameter of 9.0 mm and thickness of 1.05 mm. Measurements of two larger worms, respectively, as follows, total lengths: 30.7 mm and 40.0 mm. Thoracic widths 5.7 mm and 4.7 mm. Lengths of operculum and peduncle 10.9 mm and 9.2 mm. Lengths of the opercula alone, 5.1 mm and 4.5 mm. Distal end of operculum is not perfectly circular; hence the larger and smaller diameters in the first specimen are 5.6 mm and 5.0 mm. Abdomen in one specimen is 19.8 mm long, has 175 segments with capillaries commencing from the 137th segment; 24.6 mm long in the second, and has 155 segments, with capillaries commencing from the 124th and 60 plus the operculum on the right; it is slightly higher in the second specimen, which has 62 radioles plus the operculum on the left and 56 plus the rudimentary operculum on the right. Pinnule-free tips are very short, slightly thicker than the pinnules. Prostomial ocelli and thoracic glands absent; an apron is present. In the first specimen, the abdominal segments number 175 and capillaries start from segment 137; 155 segments in the second.
Tube. Whitish to caramel coloured and thick; three faint longitudinal ridges present in some places; four closely formed peristomes present anteriorly in first specimen (Fig. 28F).
Worm. Operculum funnel-shaped. As seen from measurements, its distal part is nearly circular, somewhat wider in a dorsoventral direction than laterally, its maximum diameter greater than length of funnel (Figs 28G, 29A). Number of radial lobes 75 in first specimen, 77 in second, bluntly triangular; cuticle thin, tubercles present along radii (Fig.  29A). Proximal part of funnel elongated, gradually decreases in diameter to where it joins the peduncle; constriction between operculum and peduncle faintly recognizable (Figs. 28G, 28A). First specimen has 58 radioles plus rudimentary operculum on right; 60 plus operculum on right; higher in second specimen, with 62 plus operculum on left, and 56 plus rudimentary operculum on right. Pinnule-free tips of radioles: very short; slightly thicker than pinnules. Prostomial ocelli and thoracic glands: absent. Apron: present.
Chaetae. Special collar chaetae bayonet shaped, and bear 2 or 3 conical processes at base of blade (Fig. 29B); may be considerably abraded in older chaetae Fig. 29D). Unserrated notch, as seen in chaetae from deep within fascicle short to moderately long, about twice as long as teeth on boss (Fig.  29B,C), long in older chaetae (Fig. 29D). Thoracic uncini: saw-shaped, with 5 teeth in single row (Fig. 29E); anterior abdominal uncini similar, with 4 or 5 teeth (Fig. 29F). Gap separating anterior abdominal tori (Fig. 28H): widest between first pair, then gradually becoming narrower till about 10th pair, where it is narrowest, after which it remains nearly the same for the next dozen segments or so, and then widens slightly in most of the abdomen.

Remarks.
As stated by ten  Serpula magna Straughan 1967 is Serpula jukesii Baird, 1865. Straughan (1967a was dealing with more than one species of Serpula under her descriptions of S. vermicularis. Straughan's (1967a) figure 3a is not of S. vermicularis. The very high number of opercular radii (80) suggests that some of them were possibly S. jukesii. MLR and a LLR along either side of it clearly recognizable in posterior part of tube (Fig. 29G).

Serpula nudiradiata n.sp.
Worm. Distal part of operculum circular, followed by a funnel-shaped part, latter like an inverted cone decreasing in diameter noticeably and having a constriction between latter and a bulbous proximal part before it joins the peduncle (Fig. 29H,I). A sharp constriction present between bulbous part of operculum and the moderately thick peduncle (Fig.  29H,I). Opercular radii terminate in 42 rounded marginal lobes; their cuticle thin and translucent; tubercles absent ( Fig.  29 I). Number of branchial radioles: 24 on each side, pinnulefree tips moderately long and slender; operculum on right, rudimentary operculum on left (Fig. 29H). Thoracic glands represented by a pair of openings on ventral side of thorax (Fig. 29K), with reddish pigmentation around them. Apron is present (Fig. 29K).
Chaetae. Eight bayonet-shaped special collar chaetae present in one of the fascicles exam ined. Each has a comparatively stout shaft, a conspicuously squarish boss which bears 2-6 triangular teeth, and a transverse row of small serrations, of variable length, either preceding them or at the angle between the boss and the blade (Figs 29L, 30A-E). An unserrated notch is absent (Figs 29L, 30A-E), even in older worn-out chaetae (Fig. 30B), a characteristic feature of present species. both H and I the thin cuticle, the proximal bulb-shaped part and distal funnel-shaped part, and a sharp constriction present between the opercular peduncle and the operculum; the operculum also has basal bulb, with a constriction between is also present between the proximal bulbous part and the distal funnel-shaped part; (J) dorsal view of abdomen showing the narrow gaps separating abdominal tori.; the tori almost meet from about the 9th/10th pairs to about the 16th/17th pairs and the gaps gradually increase again; (K) ventral view of thorax showing the 2 apertures for thoracic glands, and the apron; (L) bayonet-shaped collar chaeta, showing stout shaft, squarish boss, more or less triangular and blunt teeth, and lack of an un-serrated notch.
Thoracic uncini saw-shaped, with 5 teeth in a single row, the 3 posterior ones are considerably smaller than the 2 anterior to them (Fig. 30F,H). Anterior abdominal uncini smaller, with 4 or 5 teeth in a single row, of which the posterior 3 or 4 are abruptly smaller than most anterior tooth and approximately equal size (Fig. 30G); posterior abdominal uncini rasp shaped (Fig. 30 I). Corresponding pairs of anterior abdominal uncinal tori separated dorsally by a very short gap; widest gap between first pair, then decreases quickly from second, very narrow from about 3rd to 9th; almost meet from about 10th to the 20th (Fig. 29J); then widens, with tori being comparatively close together in rest of abdomen.
Remarks. Serpula nudiradiata differs from S. jukesii in several respects. It lacks tubercles along the opercular radii, and possesses a sharp constriction between its peduncle and the operc ulum, a bulbous swelling in the proximal part of operculum, and a pair of apertures for the thoracic glands on the ventral side of the thorax. Its bayonet-shaped special collar chaetae are also different from those of S. jukesii: they lack an unserrated notch, and possess 2-6 teeth on the conspicuously squarish boss, and a transverse row of small serrations, which precede the larger teeth on the boss (Figs 29L, 30A-E).
Etymology. The specific name nudiradiata denotes one of the characters i.e., lack of tubercles along the radii of the operculum, which distinguishes it from certain other species, such as S. jukesii, that possess them. Tube. Tube trapezoidal in cross-section and consists of a posterior attached part and an anterior unattached part (Fig. 31A,B); its unattached portion characterized by a series of peristomes; dorsomedially, peristomes are directed anteriorly (Fig. 31A). Aperture of tube surrounded by a broad rounded anteriorly directed upper lip, separated by an indentation on each side from a ventrolateral lip; the latter two separated by a mid-ventral indentation (Fig. 31B). The dorsal lip may be partially subdivided into two by a faint indentation (Fig. 31B). The trilabiate condition, and even the partially subdivided dorsal lip, may be seen in peristomes along the tube. Worm. Number of radioles 11 radioles plus rudimentary operculum on left; radioles missing on right, but base of opercular peduncle can be recognized. Light brown clusters of prostomial ocelli clusters present. Thorax has 7 chaetigers; apron present.

?Serpula species
Chaetae. Special collar chaetae bayonet-shaped; either bear two conical to triangular teeth at the base of the blade, with a smaller tooth between them (Fig. 31D,E,G,H), or only two main teeth (Fig. 31C,D). Unserrated notch is moderately long, even in a newly formed chaeta from within fascicle (Fig. 31F,H); blade about 6 times as long as unserrated notch. Thoracic uncini: saw-shaped, with 7 teeth (Fig. 31H); anterior abdominal uncini similar, with 4 or 5 teeth (Fig. 31 I).
Corresponding pairs of anterior abdominal tori almost meet along the mid-dorsal line of the abdomen.

Remarks.
Unfortunately, the operculum is missing and it not possible to decide which genus among the Serpulinae sensu stricto it belongs to. Spiraserpula Regenhardt can be eliminated because of the lack of internal tube structures and the presence of an apron. Its tube is quite unlike those of known species of Serpula, Crucigera or Hydroides from the Indo-West Pacific, although it could belong to any one of them, although the peristomes along the tube suggest that could be a species of Serpula. Future studies based on additional material would help to determine this very interesting species

Description
Measurements. Specimen from AM W202694 has a complete tube: 43.0 mm long, external width 2.7 mm. Total length of worm is 6.8 mm, thoracic width 1.0 mm; abdominal length 5.8 mm. A specimen from AM W21408 has a thoracic width of 2.0 mm; longest operculum and peduncle 4.2 mm, operculum itself 2.0 mm long in two specimens; largest opercular diameter 1.7 mm. Juvenile specimen: width of tube 0.5 mm; total length of worm 2.1 mm; width of thorax 0. 4  mm; length of operculum and peduncle 1.5 mm, length of operculum 0.5 mm, and length of abdomen 1.6 mm.
Tube. White, trapezoidal in cross-section in first specimen, with two uniformly developed smooth longitudinal ridges along its crest, with a narrow, comparatively deep longitudinal groove running between them (Figs 32A,B,  33J). Along the lateral base of each longitudinal ridge is a lateral longitudinal groove that tucks underneath the ridge and demarcates it from the rounded flank of that side. Lateral edge along each flank decreases in thickness laterally. A translucent granular overlay is present (Fig. 32B). Tube similar in second specimen, with regard to longitudinal ridges and grooves, but the median longitudinal groove is not deep where parts of the tube appear to have grown unattached to the substratum (Fig. 32H,J); the typical form, however, is recognizable in the more horizontal part (Fig. 32 I).
Juvenile tube white, shiny on inside; two smooth longitudinal ridges present along its crest, with a longitudinal groove between them; latter wide compared to size of tube (Fig. 33J); a faint granular overlay discernible along its lateral borders; internal tube structures absent.
Worm. Operculum bell-shaped ( Fig. 32C-F,K-M); its proximal part swollen and bulbous before it joins the peduncle; constriction between operculum and peduncle sharp (Fig. 32D,F,K,M). Numbers of marginal opercular lobes in the three specimens examined: 29, 23 and 25; triangular with blunt tips in juvenile specimen (Fig. 32C-F); rounded to squarish in the others (Fig. 32K-M); cuticle thin and translucent in younger specimens (Fig. 32C-F); somewhat thicker in older specimens (Fig. 32K-N). As seen in figures mentioned, tubercles present along the radii, more prominent in older specimens. Funnel-shaped opening of the distal part of the operculum leads into a hollowed cavity with its bulbous proximal part (Figs 32D,E,K,N). Highest number of radioles amongst the three specimens, 20 plus operculum on left and 18 plus rudimentary operculum on right; smallest number 12 plus operculum on right, and radiolar crown missing on left; pinnule-free tips longer and thicker than pinnules (Fig. 32N). Prostomial ocelli and apertures for thoracic glands absent; apron present. Number of abdominal segments in a complete specimen about 79; capillaries start from about segment 45.
Juvenile worm: operculum bell-shaped (Fig. 33K,M), bears 23 marginal lobes; latter nearly triangular, with smooth tips; uncertain whether a hollow is present within operculum in the early juvenile condition (Fig. 33M). Radioles number 4 on each side; a short filamentous rudimentary operculum on left, operculum on right; pinnule-free tips short and slender. Number of thoracic chaetigers 7; an apron is present. Number of abdominal segments 32 abdominal segments, capillaries start from the twenty-third.
Chaetae. Bayonet-shaped special collar chaetae ( Fig.  33A-G) usually possess two, sometimes three teeth at base of blade; in fully formed bayonets, boss as well teeth directed away from blade ( Fig. 33A-F); in newly formed bayonets from within collar fascicle, teeth directed parallel to or slightly towards the blade (Fig. 33G). Unserrated notch short to moderately long in the older bayonets (Fig. 33A,E), yet to be developed and not evident those within collar fascicle (Fig.  33G);. Thoracic uncini saw-shaped (Fig. 33H) and bear 5 or 6 teeth; anterior abdominal uncini similar (Fig. 33H) with a similar number of teeth. Abdominal neuropodia with about 13 flattened trumpet-shaped chaetae in each chaetal fascicle. Corresponding pairs of anterior abdominal uncinal tori are lateral separated dorsally by a very wide gap (Fig. 32G).
The juvenile possesses 4 bayonet-shaped special collar chaetae per fascicle, each bearing 2-4 teeth at base of blade, of which 2 may be larger than the rest; similar to the number described for adult specimens. Thoracic uncini bear 5 or 6 teeth in a single row.

Remarks. Serpula watsoni
Willey is similar to S. amplilobata and Serpula nudiradiata described earlier in that the proximal part of its operculum is bulbous and a sharp constriction occurs between the operculum and the peduncle. However, unlike in both the latter species, S. watsoni possesses a cavity extending from the funnel-shaped distal part to into the bulbous proximal part of the operculum. Serpula watsoni possesses tubercles along its radii, whereas they are absent in both S. nudiradiata and S. amplilobata. The spirally coiled tube with mutually bonded coils of S. amplilobata (Fig. 27A) distinguishes it from tube of Serpula watsoni (Figs 32A,B,H-J). For differences between their collar chaetae see relevant descriptions. Straughan's figure  (1967a, fig. 3b) of S. watsoni does not show the bulbous proximal part of the operculum that observed by Willey (1905: pl. VIII, fig. 6) and in the present account (Fig.  32C,D,F,K,M). Straughan's figure (1967a, fig. 3a) of S. vermicularis appears to be that of a juvenile S. watsoni (vide Fig. 33J-M in the present account).
The tube of the juvenile specimen, with its two longitudinal ridges along its crest (Fig. 33J), is similar to that of older specimens of S. watsoni (Fig. 32A,B,I). Bayonet-shape collar chaetae are also similar. It is uncertain whether a hollow is present in the bulbous proximal part of the operculum in the early juvenile condition. Regenhardt, 1961 Type species. Spiraserpula spiraserpula Regenhardt, 1961. Diagnosis. Operculum chitinous; funnel-shaped, usually bearing dichotomous radii ending in marginal lobes, only a few or lacking in certain species; inserted in position of second branchial radiole; peduncle non-pinnulate; interradiolar membranes absent. Number of thoracic chaetigers: seven or more. Apron: absent. Special collar chaetae usually bayonet shaped, bearing two to a number of teeth on boss. Most anterior uncinal tooth: larger than the rest. Distal ends of abdominal neurochaetae shaped like asymmetrically 'flattened trumpets', with one side longer than other, and bear numerous serrations or denticulations. Apron absent, as an adaptation to presence of internal tube structures. Tube: internal tube structures present.

Description
Measurements. Maximum external diameter of tube 1.5 mm in places, elsewhere about 1.2 mm. Worm (radioles missing): length of thorax and abdomen 18.0 mm, width of thorax 0.5 mm; length of abdomen 16.8 mm.
Tube. Has an overall brownish mustard colour, with a lateral longitudinal pigment band on each side and a less distinct and narrower median longitudinal band. In addition to the mustard colour, a pinkish tinge present dorsally. A granular overlay is present (Fig. 34A,B), granules being larger along the lateral borders. Forwardly directed peristomes may be present (Fig. 34A,E). Internal tube structures present, consisting only of a ventral longitudinal ridge; it is somewhat T-shaped in cross-section posteriorly (Fig. 34C). More anteriorly, as the ventral ridge decreases in height, it develops a shallow longitudinal groove, which gives it a somewhat Y-shaped appearance (Fig. 34E) and, still more anteriorly, becomes wedge-shaped and skewed (Fig. 34D), then tapers to a simple low ridge. Worm. Thorax has a pair of brownish-black clusters of ocelli (Fig. 35A,C); 7 chaetigers; lacks an apron; possesses a deep ventral abdominal longitudinal groove (Fig. 35A-D) to accommodate ventral internal tube structures (Fig. 34C-E).
Chaetae. Special collar chaetae bayonet-shaped, possess a long serrated blade and several teeth on the basal boss, but lack an unserrated notch ( Fig. 35E-G). Thoracic uncini sawshaped (Fig. 35H), with 4 or 5 teeth in a single row; anterior and mid-abdominal uncini different in shape, but possess a similar number of teeth (Fig. 35I,J, respectively); posterior abdominal uncini rasp-shaped. Anterior and middle abdominal segments possess 2 or 3 flattened trumpet-shaped chaetae per fascicle. Anterior abdominal tori are short and laterally located up to about the eleventh chaetiger, then gradually become dorsolateral and longer, after which they almost meet mid-dorsally, being separated by very short distance, from about the twenty-fourth pair onwards (Fig. 35A-D).
Remarks. The single specimen available is significant ly larger than those described by Pillai and ten  and provides additional data on intra-specific size and meristic variability. In the material described in 1994, external tube diameter did not exceed 0.6 mm; maximum total length of worms was about 12.5 mm and maximum abdominal length 11.2 mm. The number of thoracic chaetigers in the present specimen, 7/7, falls within the range encountered in 1994. However, its thoracic membranes extend further backwards, to thoracic chaetiger 6/6. Maximum number of abdominal segments encountered in the 1994 specimens was only 48 for an abdominal length of 11.2 mm. It is very much higher in the Kimberleys specimens, 158 for an abdominal length of 16.8 mm. Diagnosis (emended): Operculum calcareous; depending on the taxon, opercular plate may be circular or oval, simple and flat or corrugated, concave, convex or conical, with or without branched spiny processes arising from it, or may consist of stacked disks; peduncle winged; wings may or may not be fringed. Operculum inserted in position of second dorsal radiole of left side, extremely rarely on the right; rudimentary operculum absent on opposite side. Branchial radioles may be arranged in a circle or a spiral, or may be in a circle in juveniles and spiral in adults. Inter-radiolar membranes present; may or may not be fringed. Thoracic membranes present; apron present. Thorax consists of seven chaetigers. Depending on the taxon, collar chaetal fascicles may be present in both juveniles and adults, or may be reduced or absent in adults but present in some juveniles. Special collar chaetae: may be bayonet-shaped, with a squarish serrated boss; latter may be reduced to various extents, even completely, when they may be similar to the simple bladed chaetae hat accompany them; distal part of chaetal shaft preceding boss serrated in both well-developed and reduced bayonet-shaped special collar chaetae; an un-serrated notch absent between the serrated boss and blade. Serrations of blade may be fine, short and hair-like in some taxa, longer and wavy in others. Thoracic uncini saw-shaped; their anterior process gouged. Abdominal tori lateral; uncini similar to those of the thorax, usually with fewer teeth; posterior uncini may be rasp-shaped. Abdominal neurochaetae have flattened trumpet-shaped ends, with one side drawn into a long tapering process; possess long shafts that protrude conspicuously from the abdominal wall; although may not be evident in adults of certain taxa, they are present in juveniles. The above emended diagnosis of Spirobranchus takes into account the side on which the operculum occurs. It has been reported as being always a modification of the second radiole of the left side in almost all its species. It would, however, be useful to note that, out of a hundred of the Mediterranean Pomatostegus polytrema (Philippi, 1844) (= Spirobranchus polytrema in the present paper) examined by Zibrowius (1964), only one had the operculum arising on the right. According to Imajima and ten  it may arise from the left or right in S. decoratus Imajima, 1982. Although this difficult genus has received much attention in recent decades, the taxonomy of its species is yet to be satisfactorily resolved. The Kimberleys collection includes species of Spirobranchus that, due to lack of reliable characters for separating what appear to be more than a single species, have recently been referred to in publications as belonging to Spirobranchus "species groups" or "species complexes", which need to be resolved in due course. With this end in view, the Kimberleys collections of Spirobranchus were compared with other Indo-West Pacific collections belonging to the genus available in the Natural History Museum, London, and the Australian Museum, Sydney. Although the scope of this study is not intended to be a revision of Spirobranchus and its species complexes, it has yielded interesting results, which are presented below.

Remarks
Ten Hove (1970a), in a revision of the genus Spirobranchus Blainville, 1818, recognizes the following. Firstly, that there are two main groups of species: those in which the branchial radioles are spirally arranged and the opercular wings are fringeless, as in S. giganteus, and those in which they are arranged in a circle and the opercular wings are fringed, as in S. tetraceros. Secondly, that. Spirobranchus giganteus giganteus (Pallas, 1766) occurs in the tropical and subtropical coasts of America, while S. giganteus corniculatus (Grube, 1862), sensu ten Hove 1970, occurs from the Red Sea to the central Pacific. Thirdly, that S. tetraceros (Schmarda, 1861), S. giganteus giganteus (Pallas, 1766), S. polycerus (Schmarda, 1861) and S. polycerus (Schmarda, 1861) (Schmarda, 1861).
Furthermore, ten Hove (1994) recognizes S. paumotanus (Chamberlin, 1919), which he earlier (1970) synonymized with S. giganteus, and the following species from the Indo-West Pacific: S. coronatus Straughan, 1967, S. corrugatus Straughan, 1967and S. gardineri Pixell, 1913, while expressing the possibility that the latter may consist of two species. Ten Hove & Nishi (1996), while providing a very good re-description of Spirobranchus corrugatus Straughan, 1967, show that it has an Indo-West Pacific distribution, although originally described from Australia. According to Smith (1985, unpublished thesis) what had been identified in the past as S. giganteus and S. tetraceros are each a group of closely related species, the limits of which are as yet unclear, and arrives at the following conclusions. 1. Tube colour, besides its form need to be given more importance than in the past. 2. There are three main species complexes in the Australian material studied by him, namely, those of S. giganteus, S. tetraceros and S. latiscapus. 3. The Spirobranchus giganteus (Pallas, 1766)-complex can be assigned to four species, and that the Spirobranchus tetraceros (Schmarda, 1861) complex can be assigned to three species. 4. Spirobranchus decoratus Imajima, 1982 andS. gardineri Pixell, 1913 are valid taxa, with two variant types of the latter. Fifthly, that Spirobranchus giganteus giganteus, and S. polycerus among the Caribbean species identified by ten Hove (1970), the seven species worldwide listed by the latter, and S. spinosus Moore 1923 from the Californian region are valid taxa.
Frank and ten Hove (1992) and ten Hove (1994) agree with Smith (1985), and believe that the Spirobranchus giganteus and S. tetraceros categories are in need of further revision. This includes species that had been incorrectly synonymized with others by ten Hove (1970). Under the Spirobranchus tetraceros-complex, for instance, ten Hove (1994: 113) states: "in 1970, I united various nominal taxa from circumtropical origin in a single species: S. tetraceros. Nowadays I realize that this has been an oversimplification, the taxon tetraceros probably contains a number of species… " An important character pointed out by ten  between the Spirobranchus giganteus and S. tetraceros categories relates to the opercular plate. It is oval, and arises obliquely from the peduncle in the former, whereas it is circular and arises at right angles to the peduncle in the latter. Moreover, the arrangement of the branchial radioles is whorled in the Spirobranchus giganteus category but circular in S. tetraceros category, Ten Hove (1994) also refers to his earlier (1970) S. giganteus corniculatus complex as the S. corniculatus complex.
For a recent view on this see Kupriyanova et al. (2001: 74-75). The present study shows that not only certain "species complexes" of the genus Spirobranchus described in recent years consist of more than one species, but also that certain species previously synonymized with others are distinct species.
Spirobranchus is mainly a tropical to sub-tropical genus. However, its distribution will now extend to temperate waters with, for example, Spirobranchus triqueter (Philippi, 1844) (= Pomatoceros triqueter) in the Mediterranean, Northwest Atlantic and North Sea, to the Tasmanian Sea from where a new species Spirobranchus tenhovi, is described below. Zibrowius (1968: 161) discusses the close similarities between Pomatoceros, Spirobranchus, and Pomatoleios and points out the following: the only difference between them is in their collar chaetae, which are reduced or sometimes even absent in Pomatoceros. There is no reason to retain the genus Pomatoleios, based solely on the absence of collar chaetae, since there is a tendency towards their reduction and disappearance. Even if there are populations of Pomatoleios in which all the individuals lack collar chaetae there is no justification for retaining this genus. The tendency towards reduction of collar chaetae is equally present in the genus Spirobranchus, as exemplified by S. eitzeni Augener 1918. Special collar chaetae occasionally are absent in Pomatoceros and Spirobranchus, and present in some juvenile Pomatoleios. The distinction between hooded chaetae and "Spirobranchus-chaetae" sometimes is very hard to make. However, a (needed) revision of the complex of genera falls outside the scope of this paper."

Synonymy of the genera Spirobranchus, Pomatoceros and Pomatoleios
The operculum in all three genera is a calcareous plate at the end of a winged peduncle. In Spirobranchus, it may lack processes, as in S. maldivensis Pixell 1913, or may bear processes, as in its various species, including "complexes". In Pomatoceros triqueter (Linnaeus, 1767), the opercular plate bears processes similar to those of species of Spirobranchus or may be devoid of processes, as in S. maldivensis (Fig. 52C-E). As remarked earlier, the shape of the opercular plate is oval in the S. giganteus and corniculatus category, but circular in Pillai Gravier, 1905 andConopomatus Pillai, 1960), Pomatoceros Philippi, 1844 andPomatoleios Pixell, 1913, among others, with regard to many important characters. They include the uncini, chaetae (bearing in mind the discussion on the collar chaetae), winged opercular peduncles, the calcareous opercular plates, possession of an apron, and the characteristic blue or violet body pigmentation, which is present in Pomatoceros, Pomatoleios and Spirobranchus. He concludes that they form a natural group and that, eventually, need to be reclassified more closely.
Ten Hove & Nishi (1996) state that "the differences between the nominal genera Pomatoceros, Pomatoleios and Spirobranchus are very small (if indeed existing), mainly lying in the collar chaetae. Absent in Pomatoleios, hooded (limbate) in Pomatoceros and bayonet-like with numerous hair-like teeth in the basal boss in Spirobranchus." The three genera cannot also be satisfactorily separated on the basis that special collar chaetae in Spirobranchus are bayonet-shaped and possess a squarish boss, while those in Pomatoceros are capillary-like or capillaries, and that they are absent in Pomatoleios. It will be seen from descriptions of species of Spirobranchus that follow, that there is, in the various taxa, a complete range of special collar chaetae, from those in which the boss is prominently squarish to those in which it is reduced to various extents, even to capillaries. Among them is the new species Spirobranchus zibrowii, in which the special collar chaetae, like those of Pomatoceros, do not possess a recognizable boss, but are cap illaries. In Pomatoceros, the reduction of the basal boss of collar chaetae is complete, making them capillaries, except for a very narrow blade that is discernible under high power magnification. In the light of observations of this character in the various species studied below (Fig. 36), it is impossible to separate the nominal genus Pomatoceros from Spirobranchus.  Pixell (1913) erected the genus Pomatoleios, consisting of the single known species, Pomatoleios kraussii (Baird, 1864), and distinguishing it from Spiro branchus on the basis that the former lacks collar chaetae while the latter possesses them. However, it had been shown by Zibrowius (1968), ten Hove & Nishi (1996), and confirmed in the present study, that although collar chaetae are absent in adults, they are present in some juveniles. Hence, the use of absence of collar chaetae in Pomatoleios as a character to distinguish it from Spirobranchus would lead to an odd taxonomic situation in which juvenile Pomatoleios having collar chaetae are assignable to Spirobranchus and adults without them to Pomatoleios.
The three nominal genera cannot also be separated on the basis of their radiolar arrangement i.e., that it is spiral in the Spirobranchus giganteus and Spirobranchus corniculatus complexes, but circular in the Spirobranchus tetraceroscomplex, Pomatoceros and Pomatoleios. This is because, as described by ten Hove & Nishi (1996: 89) in Spirobranchus corrugatus Straughan 1967, for example, the radioles occur "in two perfect circles in small specimens," but in "1½ whorls in large specimens". There is also S. paumotanus (Chamberlin, 1919) dealt with later in this account, in which radioles are shown to be arranged in a circle in juveniles, but with up to 5 spirals in adults.
In a cladistic analysis Pillai (2008: 45) showed that according to criteria previously employed to separate the three genera, they belong to a cohesive group in which Pomatoleios is sister to the clade consisting of Pomatoceros & Spirobranchus. As a follow-up in the present paper, however, Pomatoceros and Pomatoleios are synonymized with Spirobranchus for the foregoing reasons.

Terminology relating to collar chaetae in Spirobranchus
The terminology hitherto used to describe collar chaetae of Spiro branchus poses some problems. As discussed earlier, the collar fascicle in serpulimorph polychaetes consists of two kinds of chaetae, special collar chaetae which are modified in various ways and are important in the taxonomy of its species, as well as those that end distally in simple blades. Where the special collar chaetae possess a recognizable boss and a blade, they have been termed bayonet-shaped or Spirobranchus-type. However, the latter descriptive term becomes inapplicable for the genus Spirobranchus as a whole since there is a complete range from bayonet-shaped to simple blades in the various species, as shown in this paper.
The second type of collar chaetae as well as those of the remaining thoracic segments have been termed "hooded chaetae", among others, such as simple limbate chaetae, simple bladed chaetae and bladed capillaries. However, if we are to define chaetae in relation to the Polychaeta as a whole, truly hooded chaetae, aptly termed "soies encapuchonnees" in the French literature, are found in certain other polychaete groups, such as the Spionidae and the Capitellidae, but absent in serpulimorphs. As to be expected, the "hoods" of hooded chaetae have an opening on one side in the former taxa, unlike in those described as hooded chaetae in sabellids and serpulimorphs. The terminology of serpulimorph chaetae is under review (ten Hove, personal comm.).
Abdominal neurochaetae of Spirobranchus have been described as having flattened trumpet-shaped ends with one side drawn into a long tapering process. However, these 'trumpet-shaped' ends gradually become reduced to such an extent that although they appear to be capillaries towards the posterior end of the abdomen, they are actually narrowly bladed at their tips, as also seen in many other serpulid genera.

Description
Measurements. Total length 16.6 mm; width of thorax 2.5 mm; length of operculum and peduncle 6.0 mm; length of shortest opercular horn (occurring dorsally) 2.0 mm, longest horn (occurring ventrally) 2.3 mm; length of abdomen 7.9 mm; number of abdominal segments 65.

Tube. Missing.
Worm. Radioles colourless and translucent, thorax and abdomen opaque and colourless; opercular plate white, with a translucent rim, its horns white with transparent medial and lateral spines; operculum on left, rudimentary operculum absent; opercular plate circular, bearing a circle of 8 very long horns and a short central process ( Fig.  37A-C). Each long horn bears a row of somewhat long usually 3, sometimes 4, pointed medial spines, and short lateral spines (Fig. 37C); medial spines unbranched, those towards the base comparatively longer and hooked towards the centre; lateral spines may bear 2 or 3 smaller spines. Peduncle winged, wings unfringed (Fig. 37A,B). Radioles: 23 on each side, arranged in a circle, their pinnule-free tips short and thick, pinnules very fine. Inter-radiolar membranes unfringed. Collar large, its ventral lobe triangular, lateral lobes rounded and reach third or 4th thoracic chaetiger; 7 thoracic chaetigers; apron present. Remarks. Spirobranchus baileybrockae n.sp. is identical with regard to all the important characters of the worm described by Bailey-brock (1985: 204, fig. 8a-f) from Fiji as variant forms of Spirobranchus tetraceros (Schmarda, 1861). Although the latter were not examined in this study, Bailey-Brock's very good descriptions of the two Fijian specimensare adequate for the discussion and conclusions which follow. Firstly, although the tube of the Kimberleys holotype is missing, Bailey-Brock's description of it in two specimens is in contrast with that of S. tetraceros. In the larger of the two, tube colour is "salmon-pink with a faint rose tinge to the anterior part of median ridge"; in the smaller specimen, the earlier formed part of the tube is white, "vivid rose-pink" along the median longitudinal ridge and extending anteriorly towards the aperture. As regards tube shape, that of the larger specimen is triangular in cross-section, has broad lateral flanges, and a well-formed sinuous median longitudinal ridge which projects conspicuously over the aperture, and its surface is pitted on the sides and flanges giving it a somewhat "foraminous appearance". On the other hand, tube-colour in S. tetraceros (Schmarda, 1861) as observed in the Kimberleys collection is white in adults, may be white to bluish-grey in juveniles, but not pink or rose, and does not have a pitted surface and a "foraminous appearance".
Secondly, the number and form of the main opercular horns in the Kimberleys holotype agree fully with Bailey-Brock's Fijian specimens. In both they consist of a circle of 8 long unbranched horns, which are not dichotomously branched, each bearing a regular single row of medial spines, as well as a short dichotomously-branched spine arising from the centre of the latter circle of long spines. In contrast, the number of main horns in S. tetraceros from the Kimberleys is 3, and they are dichotomously branched. Furthermore, inter-radiolar membranes are not fringed in S. baileybrockae but fringed in S. tetraceros.
Another species that needs to be compared with S.
baileybrockae is S. coronatus Straughan (1967). It has been described as possessing 8-10 main opercular horns. However, differences in their form distinguish it from the present species. While the opercular horns in S. baileybrockae and S. tetraceros are directed anteriorly (more or less perpendicularly to the opercular plate), those in S. coronatus are directed radially (more or less flattened along the plane of the opercular plate), and bear numerous lateral processes. While the 8 long spines in S. baileybrockae are not dichotomously branched, they are dichotomously branched in S. tetraceros and S. coronatus. Moreover, S. coronatus lacks the row of 3 or 4 pointed medial spines along each main spine, as well as the short dichotomously branched horn arising from the centre of the circle of main horns, which are characteristic of S. baileybrockae (Fig. 37C). Furthermore, inter-radiolar membranes are unfringed in S. baileybrockae, but fringed in S. coronatus and S. tetraceros.

Description
Measurements. Only one out of the 5 specimens in AM W202960 has an undamaged tube; its maximum external diameter, 4.7 mm. Specimens in AM W202959 are larger; maximum external tube diameter of 11.5 mm and tube thickness of 2.0 mm; however, the worm is incomplete. The largest worm has a total length of 43.8 mm, and about 217 abdominal segments. The largest thoracic width of 7.7 was encountered in the incomplete specimen from AM W202958. The largest operculum, together with its peduncle and horns is 13.5 mm long, horns alone 3.2 mm. Long and short diameters of the opercular plate in an older specimen, AM W202959, 7.5 mm and 6.5 mm, respectively.
Tube. Tube colour ranges from yellowish orange in specimens from AM W202960, to bright orange, both externally and internally, and in the very large incomplete specimen from AM W202958. Tube of small complete specimen from AM W202960 has a single, somewhat smooth, median longitudinal ridge, with fine transverse ridges on either side of it (Fig. 38A).
Worm. Colour: opercular plate rose coloured in four out of five specimens in AM W202960 white in fifth; white, and its rim and horns bright pink in one specimen; bright pink, especially its rim, and tips of opercular horns, in another, from AM W202952.
Operculum occurs on left side in all specimens, rudimentary operculum absent; opercular plate oval, the broader part being anteroventral, the whole having a posterodorsal slope (Fig. 38A-D, 39A,C,D,I,J, 40A-C). As also seen in latter figures, a pair of horns, borne on a short common base arises posterodorsally from it; each horn has a short bifid dorsomedial spine towards its base, a short posterolateral spine towards its middle, and ends in a bifid tip. Tips of main horns and spines are wasted through abrasion against the tube, the operculum evidently being withdrawn obliquely into the tube, unlike in other species of Spirobranchus. Peduncular wings unfringed (Figs 38A-C, 39A,B, 40A-C). Variations observed in arrangement of spines on the opercular horns: in an older specimen from AM W202959, tips of the main horns curved, pointed, and directed anteroventrally (Fig. 39B-D), as also figured by Mörch (1863: fig. 26). In addition, a small, anteriorly directed conical swelling may be present at the base of the short column from which the main horns arise (Fig.  39C), apparently uncommon; latter is lacking in specimen shown in (Figs 38C, 39I, 40C). A long medial process may occasionally be present in place of the conical process (Fig.  40A,B). Radioles numerous and arranged spirally on each side. Four out of the five specimens in AM W202960 possess 7 spirals on each side, the fifth 5. However, 10 spirals are present on each side in the specimen from AM W202958. A broad membrane extends between the radiolar whorls. Translucent ocelli-like structures present laterally along the radioles and at the bases of their pinnule-free tips. Interradiolar membranes unfringed (Fig. 38A). Highest number of abdominal segments counted, 226 in a worm from AM W202959. Long-shafted neurochaetae and notopodial uncinal tori are present in all except the first 3-6 or so abdominal segments.
Chaetae. In bayonet-shaped special collar chaetae (Fig.  39E), a serrated area of the chaetal shaft precedes boss; latter squarish, prominent; unserrated notch absent between boss and blade; serrations along distal part of blade elongated and wavy, those of proximal part much shorter; blade about twice as long as combined length of the serrated part of chaetal shaft plus boss. Uncini saw-shaped, anterior process gouged; thoracic uncini ( Fig. 39F) very long compared to those of the other species of Spirobranchus, and with a higher number of teeth, around 22 to 24; anterior abdominal uncini much smaller, with about half the number of teeth (Fig. 39G); posterior abdominal uncini similar, with about 13 teeth (Fig. 39H).

Remarks. Spirobranchus corniculatus is a widespread
Indo-Pacific species (see ten Hove 1970: 14). Material described as S. giganteus (Pallas) by Straughan (1967: 243) from Queensland, with its characteristic oval operculum and structure of opercular horns (Straughan, 1967: fig. 14e), belong to S. corniculatus. Likewise, specimen BMNH 1938.7.25.2, from Hulule, Male Atoll, Maldives, labelled type specimen, and as a new variety Spirobranchus giganteus var. turbinatus, by J.S. Gardiner, which was examined during the present study, also is S. corniculatus. It is very large, and has a total length of 90.3 mm; its operculum is shown in Fig.  40C. However, its radioles are arranged in only 6 spirals on each side, compared to the highest number recorded during this study. The number of abdominal segments is 226. The largest worm from Kimberleys AM W202960, however, has a total length of 43.8 mm, half that of the above specimen from the Maldives, and about 217 abdominal segments. The highest number of abdominal segments observed is 226 in a worm from Kimberleys collection AM W202959. The highest number of branchial whorls is 10 on either side in a specimen from AM W202958. Its special collar chaetae bear serrations, which are, as described earlier, long and wavy. Its thoracic uncini bear up to about 21 teeth, in addition to the anterior gouged process; and anterior abdominal uncini, about 11 teeth.
In a larger specimen of S. corniculatus in BMNH 1970.853 from Matur Island, Solomon Islands, collected and presented to the NHM by P.E. Gibbs in 1970, the radioles are arranged in 8 spirals on each side. Opercular peduncle and inter-radiolar membranes are unfringed. Abdominal uncinal tori are lateral, starting from about 7 or 8 segments from the anterior end; long-shafted abdominal chaetae present throughout the abdomen.
In a very much larger specimen, in BMNH 1970.82, collected by P. Gibbs from Paleki Islands, LWM, during the Royal Society Expedition to the Solomon Islands, and identified as S. giganteus corniculatus by ten Hove,7.9.95, of total length 99.5 mm, has only 8 radiolar spirals on each side.
Although S. bucephalus (Mörch, 1863), based on Pomatoceros bucephalus Mörch (1863: fig. 26) from Java, and accompanied by a more accurate description.existed asan alternative for S. corniculatus, Grube's description came out earlier, in 1862, and has priority in terms of the rules of Nomenclature See ten Hove (1970) for past records of S. corniculatus from the Philippines, Indonesia, New Guinea, New Caledonia, Marshall Islands, Australia, Japan, Maldives, Sri Lanka, S. Arabian Coast, Red Sea, East Africa, Madagascar and South Africa.

Description
Measurements. Total lengths 7.0 mm and 6.7 mm; widths of thorax 1.0 and 0.9 mm; lengths of operculum and peduncle 3.0 mm and 2.7 mm; abdomen incomplete in first specimen, 2.8 mm long in second. Number of abdominal segments in complete specimen: 41, with capillaries commencing from the 33rd.

Tube. Missing.
Worm. Operculum calcareous; occurs on left side in both specimens, rudimentary operculum absent. Opercular plate bears a pair of rounded lobes (Fig. 40D-F) similar to that described for S. polytremus (Philippi, 1844); peduncle winged, the latter fringed (Fig. 40F). Branchial radioles arranged in a circle on each side; 7 pairs of radioles in first specimen; 6 on the left and 7 on the right in second; pinnulefree tips moderately long and thick. Thoracic chaetigers number 7. Apron: present.
From various past descriptions from the Indo-Pacific area, it appears that there are a number of Spirobranchus polytremus-like species, which need to be sorted out. A good example is that of the types A and B of this complex described by Imajima (1977: 102-106). Imajima remarks that "It is unlikely that" the species described by him "will prove to be the same as the Mediterranean species". However, regarding the "complexity of the group involved, the confusion of names by previous authors, and the scarcity of data on infraspecific variation of the individual species, it appears best not to introduce another name, and await a full-scale revision of the group." For the same reason, the Kimberleys specimens are therefore not named.
It would be useful, however, to note the differences among the Mediterranean S. polytremus, the similar Kimberleys specimens described above, and other similar species described from the Indo-West Pacific. According to Zibrowius (1968), the tube of S. polytremus (Philippi, 1844) from the Mediterranean possesses a well-developed MLR and a pair of LLR's on either side of it, with rows of pits at their bases. According to Bianchi (1981: fig. 41A,B) its colour is white, but may be faintly rose or violet. The description of the tubes by Ben-Eliahu & Fiege (1996) agrees with that of both Zibrowius and Bianchi: "Characteristic tubes triangular in cross section with tiny pits at medial and lateral keels (some with faint pink cast), and with typical internal cellular chambers within the lower lateral parts…" The worm in S. polytremus, according to Zibrowius (1968), possesses about 9 radioles, the operculum occurs in the position of the most dorsal radiole of the left side, the peduncle is winged and the wings are not fringed. In the Kimberleys described above are different in that their peduncular wings are fringed (Fig. 39D-F); and the maximum number of radioles is 6, but confirmation is required as to whether there are larger sizes with a higher number of radioles or not.
Among the similar species from the Indo-West Pacific: Temporaria polytrema from New South Wales (Straughan, 1967) and Sri Lanka by (Pillai, 1971), the synonymies of which had previously been dealt with, and specimens of a very small S. polytremus-like species collected by the present author from Jayapura, West Irian (Indonesia) in 1969/70 and passed on to ten Hove. Spirobranchus oligotremus (Straughan, 1967) is another. Straughan (1967) states that the former has about 8 pairs of radioles arranged in a spiral and inter-radiolar membrane are absent. It is unusual for a Spirobranchus to have such a low number of radioles spirally arranged on each side. In other S. polytremus-like species, however, the radioles are arranged in a circle on each side and inter-radiolar membranes are present.

Description
Measurements. All tubes found growing on and covered over by coral. Holotype, AM W21433: tube diameter, 6.5 mm; total length of complete worm in latter 42.4 mm, its thoracic width, 3.5 mm; length of operculum and peduncle 10.7 mm, and the opercular horns 3.5 mm; abdomen 33.7 mm long, and the number of abdominal segments about 145. Largest tube diameter, 7.0 mm, occurs in one of the paratypes in AM W21433. Total lengths of specimens from AM W21442 range from 9.5 mm to 40.9 mm, the latter almost the same size as the largest from AM W21433. Smallest specimen, from AM W21442, has a total length of 9.5 mm. However, the early juvenile in AM W21474 has a maximum external tube diameter of 0.5 mm; total length of worm 4.4 mm, its thorax 0.4 mm wide, length of operculum and peduncle 1.7 mm, length of operculum alone, 0.5 mm, and the length of abdomen 1.7 mm.
Tube. Tube colour of holotype: mauve patches on a white background; paratypes: one has a darker mauve coloration than others; overall colour range of tubes in all material studied yellow to yellowish brown, and light to dark mauve. Holotype shows a low median longitudinal ridge and 4 or 5 narrower ones on either side of it (Fig. 41A), while a paratype is similar, but its lateral ridges are less clearly developed (Fig.  41J); fine transverse ridges present throughout (Fig. 41A,J).
Worm. Opercular plate in holotype white; pinkish to bright pink in older specimens; horns white, pinkish to purplish in places. Operculum occurs on left in all specimens; no rudimentary operculum on right. Based on several specimens of varying sizes, operculum circular in juveniles and intermediate sizes (Figs 41E, 42A,C,E,F); somewhat oval in some older specimens (Fig. 41M,N). Arising vertically from opercular plate in holotype is a column bearing a pair of dorsolateral horns and a bifid ventromedian horn ( Fig.  41B-E); tips of ventromedian horn curved towards opercular plate. Dorsolateral horns not dichotomously branched ( Fig.  41B-E); each ends in an anteroventrally curved, more or less pointed tip, and bears three short accessory spines, one located ventrolaterally towards its middle or at the point where it joins the central column, and the other two located sub-terminally; paratypes and other specimens similar (e.g., Figs 41K-N, 42A-E). Central column itself may be short and stout as in holotype (Fig. 41B-E) and other specimens (e.g., Fig. 42A,B), or comparatively longer and less stout (Fig. 42E,G). Peduncular wings unfringed in holotype and paratypes (Fig. 41B,C,K), and all other specimens similar (e.g., Fig. 43A-D,K). Radioles numerous, arranged in 4½ whorls on each side in holotype, the maximum number observed; up to 3½ whorls in other specimens; smallest specimen (from AM W21442) has 19 radioles plus the operculum on the left and 25 radioles on the right, arranged in a partial spiral on both sides. Inter-radiolar membranes unfringed (Fig.  42B); pinnule-free radiolar tips as long as, and of same thickness as the pinnules. A pair of blackish clusters of prostomial ocelli present ventrolaterally, anterior to the first pair of thoracic tori (Fig. 42E). Seven thoracic chaetigers; the distances between corresponding pairs of thoracic tori decreasing toward thorax where they meet or almost meet; apron present ( Fig. 42B,D Fig. 43H for another specimen) are smaller but bear similar number of teeth. Abdominal neurochaetae with long shafts present throughout the abdomen (Fig. 42B,H).
Two specimens in the collection appear to be aberrant. Although similar to the above in having a circular opercular plate and three horns arising from a long central column, the horns themselves differ, as follows: In the smaller one (Fig.  43A), the dorsolateral pair of horns are blunt and appear abraded, while all horns are blunt in the other (Fig. 43B-D). However, bayonet-shaped special collar chaetae (Fig. 43E) are similar to those described earlier. Thoracic uncini similar, but bear a lower number, about 15 teeth in a single row, and an anterior gouged process (Fig. 43F); anterior and posterior abdominal uncini similar and bear about 10 and 9 teeth, respectively (Fig. 43G,H).
In the juvenile specimens from AM W21442: tube white, bears traces of two longitudinal ridges and faint transverse ridges (Fig. 43 I). Operculum on left side, no rudimentary operculum on right; opercular plate circular, flat (Fig. 43J,K), bears a central column at the end of which are three rudiments of soft, unbranched, tapering horns; peduncle winged; wings unfringed. Radioles number 6 per side, arranged in a circle on either side, as to be expected in early juveniles; inter-radiolar membranes present; number of thoracic chaetigers 7; abdomen consists of 25 segments; neurochaetae with long shafts protruding conspicuously from the abdominal wall present from about segment 15; abdominal uncinal tori short and lateral.
Remarks. Spirobranchus richardsmithi n.sp., is super ficially sim ilar to S. gardineri Pixell, 1913 with regard to the circular opercular plate and column arising from it bearing three processes distally. However, unlike in the present species, the three distal processes in S. gardineri are not bifurcated, and are directed anteriorly S. gardineri as described by Pixell (1913: 81, plate 8, fig. 7a-c), and as confirmed by ten  in material from the Seychelles and Amirantes, directed "away from the opercular plate". According to the latter, "forms with re-curving ventral spines occur in the Indo-West Pacific too; may be the taxon gardineri contains two species." Pixell (1913) also mentions that the three anteriorly directed processes "sometimes give rise to short secondary branches". A very important character of S. gardineri relates to the size and proportions of its thoracic uncini. As stated and figured by Pixell (1913: 81, plate 7, fig. 7c), they are very large. It is confirmed after examination of Pixell's slides (types) in the Natural History Museum during the present study that they are indeed quite unlike those of the other species dealt with in this account, in being extraordinarily long and bearing at least 25 teeth plus the anterior gouge. Bailey-Brock (1985) describes material from the Fijian coral reefs similar to S. richardsmithi from the Kimberleys, but assigned them to S. giganteus corniculatus (sensu ten Hove, 1970). Smith (unpublished thesis, 1985) describes similar forms from Queensland, in which the most ventral of the three processes at the end of the common column is bifid, which he refers to Spirobranchus gardineri, Pixell, 1913 (sensu stricto), with two variant types. However, specimens from the Seychelles and Amirantes studied by ten Hove (1994) agree with those described by Pixell (1913) in that all the spines at the distal end of the common column are directed anteriorly, i.e., away from the opercular plate. The single very large specimen described above and identified by by Bailey-Brock (1985: fig. 6a-c) therefore appears to belong to the new species, S. richardsmithi. Its opercular plate is circular, unlike in S. corniculatus, described earlier in this account, in which it is oval, and, as in S. richardsmithi, the two dorsolateral processes arising from the central column are curved anteroventrally, while and tips of its bifid ventral process are curved towards the opercular plate.
Another synonymy is that of S. gardineri Pixell, 1913, described from Hainan, South China Sea by Fiege & Sun, (1999: 126). The number of teeth in the thoracic uncini, in addition to the gouged process, in the Kimberleys material of S. richardsmithi is about 15-20, while it is 20-21 in the specimens from Hainan. As shown above, it is much higher, about 26, in S. gardineri (Pixell, 1913). Fiege & Sun also state that although the tubes are embedded in coral, with only their openings visible, they have a purplish interior, which is similar to the mauve tube colour of S. richardsmithi from the Kimberleys, so suggesting that it belongs to the latter species.
The early juvenile of S. richardsmithi described above shows a possible plesiomorphous condition in an ancestor from which S. gardineri, S. richardsmithi and an un-named species described later in this account could perhaps have been derived.
As seen from the drawings and notes kindly provided by ten Hove in 1996, the new species also occurs in Indonesia.
Etymology. Named after Dr Richard Smith who considered it to be a distinct species (1985, unpublished PhD thesis), and his interesting contributions to our knowledge of serpulimorphs. Figure 44. A-C, Spirobranchus tetraceros (Schmarda, 1861), specimen from AM W21438. (A) tube with worm in situ; tube obtusely triangular in cross-section, with a low median longitudinal ridge; (B) left dorsolateral view of anterior end of worm showing dichotomously branched opercular processes, fringed peduncular wings, fringed inter-radiolar membranes, and the thoracic membranes; (C) operculum viewed from anterior end; showing (apparently) 4 main branches, but actually 3 dichotomously branched processes arising from a common base.

Description
Measurements. Maximum external tube diameter in available specimens 4.0 mm; minimum 1.5 mm in a juvenile in AM W21441. Total length of larger worm in AM W21441, 16.1 mm; abdomen 8.9 mm long, 78 segments.

Tube.
Mostly covered over by coral; difficult to remove without damage to external ridges.
Colour in larger specimens usually white, with a light caramel coloured inner layer; both specimens from AM W202953 white, with a bluish-grey band along the median ridge, especially in their earlier formed parts; obtusely triangular in cross-section in older Figure 45. A-I, Spirobranchus tetraceros (Schmarda, 1861). A,B, a worm from AM W21441: (A) dorsal view of whole worm; (B) ventral view of anterior end of the latter worm; shows the branched horns, mottled ventral side of peduncle, fringed interbranchial membranes, pinnule-free radiolar tips, raised uncinal pads, the last two meeting mid-ventrally and the apron which is contracted into a triangular lobe, since the worm was fixed within its tube, and the long-shafted abdominal chaetae. C-F, chaetae from same specimen: (C,D) bayonet-shaped special collar chaetae; (E,F) thoracic uncini. G-I, a specimen from AM W21441: (G) dorsal view of anterior end, highly contracted, showing fringed opercular wings and fringed inter-radiolar membrane; (H) ventral view of anterior end showing inter-radiolar membranes, ventral lobe of collar, thoracic tori not meeting ventrally since the worm was evidently fixed outside its tube, the apron, and anterior abdominal chaetae with long shafts; (I) opercular spines viewed from the anterior end showing three main horns.
specimens with a single low median longitudinal ridge, and fine transverse growth markings extending laterally from it (Fig. 44A). Tube of juvenile from AM W21397 is white, with a high, wavy median ridge and a low dorsolateral ridge on either side of it (Fig. 46A,B); fine, shallow, transverse ridges present at irregular intervals, between longitudinal ridges and along flanks.
Worm. Colour (in alcohol): operculum and wings blue dorsally, and light blue with elongated blue specks ventrally; opercular plate white, its rim bright pink, as also are the tips of the horns. Although thorax and radioles in the juvenile specimen are colourless, they are bluish to purplish blue in the others. Basal two-thirds of branchial radioles, including inter-radiolar membrane, dark purplish blue; abdomen light purplish-blue anteriorly, gradually becomes colourless posteriorly. Operculum occurs on left side, no rudimentary operculum on right; opercular plate circular (Figs 44C,  45A,B,G-I, 46D,E,M,L). Distal end of operculum in a juvenile specimen is bulbous; its somewhat concave, circular plate bears apparently four, but really three, main branched horns arising from a common base, two of which are dorsolateral, and the third ventral (Figs 44B, 45G,I, 46E,L,M). Opercula of 7 specimens from AM W21449, as well as those of others examined are similar and bear three main processes. In an older specimen, dorsolateral process is branched thrice, the ventral one twice. Peduncular wings fringed with digitate processes anteriorly (Figs 44A, 45G, 46B-D,L,M). In a specimen from AM W21449 there is a small peaked process, centrally, at the base of the three horns (Fig. 46L,M). Radioles arranged in a circle on each side (Fig. 45B,H), their numbers ranging from 11 pairs in the juvenile specimen described above to 28/29 radioles in a specimen from AM W202956 with a total length of 41.0 mm; their pinnule-free tips longer than the pinnules that precede them. Inter-radiolar membranes fringed, extending distally to about 1 ⁄3 to ½ length of radioles (Figs 44B, 45B,H, 46C,L). Prostomial ocelli not discernible, but a pair of light-caramel patches sometimes seen in the corresponding positions. Thoracic chaetigers number 7, apron present (Figs 45B,H, 46C). Uncinal tori in thorax of adults occur on conspicuously raised pads with the distances between corresponding pairs decreasing posteriorly, leaving a triangular area between them; 5th and 6th pairs may meet mid-ventrally, as in specimens fixed within their tubes (Fig.  45B), or may be separated when worms are fixed outside the tube (Fig. 45H). In juveniles uncinal tori do not occur on conspicuous pads (Fig. 46C). Chaetae. Special collar chaetae bayonet-shaped; a conspicuous squarish boss present; area of shaft preceding the boss as well as the boss serrated; an unserrated notch absent between the latter and blade (Figs 45C,D, 46F,G), the latter two from a juvenile specimen. Thoracic and abdominal uncini saw-shaped, with a single row of teeth; ending anteriorly in a gouged process. Thoracic uncini bear about 10-12 teeth (Figs 45E,F, 46H), latter in a juvenile specimen; anterior abdominal uncini bear about 8 or 9 teeth (Fig. 46J); may increase to about 13 in posterior segments (Fig. 46K). Abdominal neurochaetae possess long shafts throughout the entire length of the abdomen (Fig. 45B); their flattened trumpet-shaped distal ends asymmetrical, with one side drawn into a tapering process (Fig. 46 I).
Remarks. From the distribution provided by ten Hove (1970a), it appears that a single "tetraceros-type species", S. tetraceros, occurs from the Caribbean region to the Indo-Pacific, which would also make it a very ancient species. According to ten Hove (1970: figs 7, 8), there is overlap in the opercular structure between samples from Curaçao (Spaanse Baai) and those from the Indo-West Pacific. However, Spirobranchus tetraceros was originally described from New South Wales as Pomatoceros tetraceros by Schmarda, 1861. As stated by Smith (1985, unpublished thesis) and ten , collections of Spirobranchus tetraceros identified by workers in the past appear to consist of a complex of species, and detailed studies of different populations are needed to sort out their identities. Nevertheless, the Kimberleys specimens agree with the description of Spirobranchus tetraceros (Schmarda, 1861), and are, therefore, assigned to it.
Trinomial nomenclature was used in some instances to designate specimens/populations that were different from a nominal species with regard to certain characters, as varieties of that species, e.g., S. tricornigerus Grube, 1878var. racemosus Pillai (1971 from Sri Lanka, and S. tricornigerus decoratus Imajima (1982) from Micronesia. Imajima and ten  recognized them as being identical and named them as a distinct species Spirobranchus decoratus Imajima, 1982. Both varieties have the following common characteristics: tube pink on a white background, triangular in cross-section, with a high median longitudinal ridge, and 5 longitudinal rows of pits; operculum bears three main branched processes, compact and directed parallel to the plane of the opercular disc rather than away from it. Peduncular wings fringed; branchial radioles arranged in a circle on each side; inter-radiolar membranes present. Special collar chaetae bayonet-shaped; thoracic uncini and anterior abdominal uncini have about 12 teeth.
From descriptions of the species which follow, it will be seen that the opercular horns and opercular plate in certain Spirobranchus collections/populations from different Indo-West Pacific locations are so superficially similar to those of S. tetraceros that they can, upon cursory examination, be assigned to the "Spirobranchus tetraceros-complex".
However, more detailed comparisons between samples or populations, and of character combinations, including tube form, and colour (differences in the latter, presumably having a biochemical basis), structure of opercular processes and chaetae, could reveal distinct species of Spirobranchus. For instance, Smith (unpublished thesis, 1985) describes three unnamed species belonging to the S. tetraceros group, species A, B, and C, from NSW and Queensland. Moreover, ten Hove (1994: 113) states: "in 1970, I united various nominal taxa from circumtropical origin in a single species: S. tetraceros. Nowadays I realize that this has been an oversimplification, the taxon tetraceros probably contains a number of species … " It would, therefore, appear more useful, in the long run, to acknowledge that certain collections had been unidentifiable for one reason or another, and retain them as un-named taxa, while awaiting the results of further research, rather than designate them as belonging to currently unrecognized taxonomic categories, such as "… complexes" or "… groups".

Description
Measurements. Maximum measurements follows: tube diameter 1.5 mm, total length of worm 8.7 mm, width of thorax 0.9 mm, length of operculum and peduncle 8.7 mm, operculum 1.0 mm, abdomen 0.9 mm.
Tube. Colour faintly pink. Three longitudinal ridges present dorsally, and one along each flank, as well as narrow transverse ridges (Fig. 47A). Median longitudinal ridge high and wavy; those on either side of it and along the flanks are low. Outlines of the longitudinal ridges irregular due to the presence of short digitate processes extending from them, apparently freshly overlaid with coral.
Worm. Operculum on the left side; consists of a conical, pitted, calcareous cap; distal end of latter slanted, blunt and lacks processes (Fig. 47B,C); peduncular wings fringed anteriorly. Maximum number of radioles counted 10 per side; a fringed inter-radiolar membrane present. Prostomial ocelli not discernible, thoracic chaetigers number 7, an apron present; highest number of abdominal segments among the three specimens, 55. Neurochaetae with long shafts present throughout the abdomen, the last 25-30 segments with longer and more slender chaetae.

Remarks.
The tube form and colour in the above juveniles are different from those of juvenile S. tetraceros. While those of the above juveniles are pink, those of S. tetraceros are white, possess a high sinuous median longitudinal ridge and a low one along each flank (Fig. 46A,B), but lacks the fine digitate processes seen in the present juveniles, the latter perhaps formed by new coral encrustation. The smallest available juvenile operculum of S. tetraceros is shown in Fig. 46B-E. As seen from the scales against the figures, its opercular plate and horns are quite small, but the plate is not conical as in the present juveniles. In the present state of our knowledge of the ontogeny of the various species of Spirobranchus, it is difficult to assign the present juveniles with conical opercula (Fig. 47B,C) to any particular species. Similar specimens were described by Pillai (1960: 21-22, fig. 8F-I), as a new species, Conopomatus sectoconus, under a new genus Conopomatus, which also included C. acuiconus. Ten Hove (1970) correctly synonymized Conopomatus with Spirobranchus, while considering the operculum of C. sectoconus to be similar to that he described as the juvenile condition of the S. tetraceros "complex". Further studies would be needed to confirm whether the juvenile Spirobranchus sp. described above, those described b Pillai (1970) and ten Hove (1970) occur in Spirobranchus tetraceros (Schmarda, 1861).

Section 2 Descriptions and comparisons with some collections of Spirobranchus from other locations in Australia and the Indo-West Pacific Oceans
Spirobranchus acuiconus (Pillai, 1960)

Description
Measurements. Tube in 4 fragments, white and bears a low, smooth medial longitudinal ridge and a lateral longitudinal ridge along the edge of a flattened part of tube on either side of the median longitudinal ridge (Fig. 37A). Total length of worm is 11. 4 mm; width of thorax 1.0 mm; length of operculum and peduncle 3.2 mm; length of operculum alone, 1.6 mm. Number of radioles: 12 plus operculum on left and 14 radioles on the right. Operculum has about 13 transverse annuli, superficially certain species of Vermiliopsis (Fig.  37B,C). However, it is calcareous, its peduncle winged, and wings fringed, unlike in Vermiliopsis in which the operculum is chitinous, and its peduncle wingless.
Remarks. Ten Hove (1970: 5) correctly synonymized Conopomatus sectoconus Pillai, 1960, with Spirobranchus tetraceros. However, there is no evidence as yet for Spirobranchus acuiconus (Pillai, 1960), to be synonymized with S tetraceros. A transversely annulated calcareous operculum lacking opercular horns or vestiges, as in S. acuiconus, has not been recorded in individuals belonging to populations of S. tetraceros in the Indo-Pacific area. It would, therefore, be useful to keep S. acuiconus distinct until evidence to the contrary emerges through future studies.

Description
Measurements. Based on three larger tubes, maximum external diameter, including the median longitudinal ridge, 5.6 mm; total length of longest worm, 39.0 mm; lengths of operculum and peduncle 9.0 mm, thorax 7.0 mm, abdomen 23.0 mm; number of abdominal segments 86; number of radioles, 34 on each side. Longest operculum and widest thorax in another specimen: 2.2 mm and 3.6 mm, respectively.
Tube. Overall colour, light to bright pink or pinkish red; one tube brownish mauve. Isolated or small clusters of granules of tube material occur irregularly on the surface (Fig. 48A). Anterior end of the tube may be peristome-like with short ridge-like extensions strengthening it. A high, wavy, median longitudinal ridge present, more or less continuous along its entire length in some tubes (Fig. 48A), or naturally interrupted or broken off in places in others. Very faint transverse ridges are present throughout, with alternating thin bands of light and dark pink coinciding with them and the grooves in between. Transverse ridges may be raised along the flank (Fig. 48A) to form a faint single or double longitudinal ridge, with a narrow, shallow longitudinal groove between them in the latter.

Worm.
Operculum on left side, no rudimentary operculum; adult opercular plate dome-shaped (Fig. 48C) to somewhat convex (Fig. 48B), with up to about 4 processes of variable length which may bear short lateral processes (Fig. 48B). All older specimens have a colonial coelenterate (Hydroida: Hydractiniidae) growing on the operculum (Fig. 48B). The horny anastomozing perisarc of the latter is very closely bonded with the opercular disk and horns, and is very difficult to remove without damaging the latter. Opercular wings fringed anteriorly. A juvenile operculum (Fig. 48D) has a short conical cap with a dorsally directed peak and lacks processes as well as the epizoic coelenterate. Colour of the radioles in alcohol light to dark blue; up to 34 radioles per side counted, arranged in a circle, their pinnule-free tips are moderately long and much thicker than the pinnules. Interradiolar membranes wide and pleated between radioles, but are not fringed (Fig. 48B); apron present.
Chaetae. Bayonet collar chaetae (Fig. 48E,F) similar to those usually encountered in the larger species of the genus in having a squarish boss; moderately long serrations present along blade; unserrated notch absent. Compared to most other species with similar bayonet chaetae, a long part of chaetal shaft itself is serrated distally, the latter almost equal to the length of the blade. Thorax uncini (Fig. 48G) bear 13 teeth in a single row and an anterior gouged process; anterior and posterior abdominal uncini similar (Figs 48H), bearing about 9 and 11 teeth, respectively, in addition to the anterior gouged process; long-shafted abdominal chaetae present throughout the abdomen.

Remarks.
The new species, Spirobranchus arabicus, is similar to S. tetraceros from the Kimberleys with regard to the circular opercular plate bearing branched horns, the arrangement of radioles in a circle, bayonet collar chaetae having a prominent squarish boss and thoracic uncini bearing about a dozen teeth. Ten Hove (1970: 3) synonymized the present species with S. tetraceros. However, there are important differences between them. The tube in S. arabicus is bright pink to pinkish-red. It is circular in cross-section in older specimens, and bears a high, wavy, fin-like median longitudinal ridge (Fig. 48A). Inter-radiolar membranes are unfringed. In contrast, the older tubes of the S. tetraceros from the Kimberleys are white, obtusely triangular in crosssection, and bear a very low, median longitudinal ridge (Fig.  44A). Furthermore, its inter-radiolar membranes are fringed unlike in the present species.

Tube. Missing.
Worm. Operculum on left side, no rudimentary operculum on the right; 20 radioles on each side, some on the left appear to have been in a state of regeneration. Operculum (Fig. 49B,C) ends distally in a simple, somewhat concave, calcareous plate, with a narrow transparent rim, lacks processes. Peduncular wings unfringed, although the tip of wing of left side is aberrantly bifid (Fig. 49C). Radioles arranged in a circle on each side; inter-radiolar membranes unfringed (Fig. 49A). Pinnule-free tips of radioles are slightly thicker than and about as long as the longest pinnules. Thorax with 7 chaetigers; gaps between corresponding pairs of thoracic uncinal tori decrease towards posterior end of thorax leaving an inverted triangular area between them ventrally (Fig. 49A), as in certain other species of the genus. Uncinal tori: borne on flap-like extensions of thoracic segments; apron present (Fig. 49A).
Chaetae. Special collar chaetae different from bayonetshaped chaetae often associated with Spirobranchus in lacking a squarish boss; however, a bend towards the smooth side of the shaft indicates transition from the serrated part of the chaetal shaft ends to the chaetal blade (Fig. 49D); serrations along blade continue to a short distance from its tip (Fig. 49D). Thoracic uncini saw-shaped; bear a row of about 10-11 teeth and an anterior gouged process; anterior and posterior abdominal uncini similar and bear about 9 teeth. Abdominal chaetae possess long shafts projecting beyond abdominal wall; one side of distal end drawn out into a tapered process.
Unfortunately, the tube of the single specimen of S. sp. 3 is missing; this needs further investigation during future studies. Meanwhile, it is worth noting the tube and form in the type specimens and additional material of S. maldivensis described in this account (Fig. 52B).

Description
Measurements. Maximum height of tube, including median longitudinal ridge, 4.5 mm; 3.5 mm excluding the latter; maximum width of tube 4.7 mm. Worm with posterior part of its abdomen broken off; posterior ends of two worms present in vial. Length of operculum and peduncle 4.7 mm, length of operculum 2.2 mm, diameter of opercular plate 2.5 mm. Radioles number 19 plus operculum on the left, and 19 on the right.
Tube. White, wide anteriorly and narrows abruptly towards posterior end (Fig. 50A,B); trapezoidal in cross-sectional outline, has a high, wavy, MLR (Fig. 50B), which is thick when viewed from above (Fig. 50A), projects conspicuously over aperture (Fig. 50A,B). Although apparently nearly flat on either side of MLR, actually gently sloping laterally (Fig.  50B); along on each side of the lateral edge of this almost flattened area, is an LLR. Smooth faint transversely arranged grooves and ridges present between median and lateral longitudinal ridge of each side (Fig. 50A,B). Longitudinal rows of foramina: absent.
Worm. Operculum on left side, rudimentary operculum absent; opercular shape conical; its apex curved dorsally and bears vestiges of three opercular horns (Fig. 50C,D); latter dichotomously branched and ending in a few short transparent spines. Granular deposits of calcareous material present along sides of operculum are (Fig. 50C). Peduncular wings fringed (Fig. 50D). Branchial radioles arranged in a circle on each side; their pinnule-free tips conspicuous and about 2½ times as long as the pinnules. Thorax consists of 7 chaetigers.
Chaetae. Special collar chaetae characterized with following: boss square and papillate (Fig. 50E,F). Serrations present all around distal part of the chaetal shaft (i.e., dorsally, laterally and ventrally), unlike in the other species of Spirobranchus dealt with in this account, where they are present only along the side of the shaft opposite to that of the blade and partly along sides of the latter. Ratio of length of blade to that of the serrated part of chaetal shaft plus boss is about 1.27: 1.0. Thoracic and anterior abdominal uncini: saw-shaped, former bearing 8-11 teeth, and latter 8, in addition to the anterior gouged process.
Abdominal chaetae possessing conspicuous long shafts projecting from the body wall present from the beginning of the abdomen.
Ten Hove (1970: 6, figs 1, 2, 15 & 16) also describes a specimen from Bahrein with a conical operculum and vestigial horns. Its tube is reddish in colour, triangular in cross-section, and possesses a prominent ridge, which projects over the aperture. However, S. elatensis differs from the latter and S. tetraceros reported from the Indo-Pacific with regard to the following characters. Its tube is white, trapezoidal in cross-section and bears a high, thick, wavy median longitudinal ridge, the latter with fin-like sections at intervals and projecting forwards over the aperture. It also bears a pair of lateral longitudinal ridges, each along the edge of a flattened part of the tube on either side of the median longitudinal ridge. Furthermore, the special collar chaetae of S. elatensis are quite characteristic, in possessing a square, papillate boss, and the distal part of its chaetal shaft being conspicuously serrated all round (Fig. 50E,F). Indeed, they are quite unlike those of the other species of Spirobranchus dealt with in this account as well as in the specimen from Bahrein described by ten Hove (1970: fig. 18).
Spirobranchus tetraceros differs from S. elatensis as follows. Its tube is white to cream or bluish, its adult tubes are obtusely triangular in cross-section and its thin, low median longitudinal ridge is comparatively inconspicuous (Fig. 44A). Its opercular horns are comparatively long and well developed. The boss of its special collar chaetae is not as square and papillate all around the distal part of the chaetal shaft, as in S. elatensis, but serrated only ventrally and ventrolaterally (Fig. 45C,D).
Etymology. Named after the type locality, Elat, Gulf of Aqaba, Red Sea.
Worm. Colour of anterior part of body, especially operculum, radioles and thorax, blue in in alcohol-preserved material. Opercular insertion: in position of second radiole of left side; opercular plate circular, with a non-transparent rim, and lacks processes (Fig. 49E,F). Radioles arranged in a circle on each side; inter-radiolar membranes present; apron present.
Chaetae. Collar chaetal fascicles: absent in adults. A juvenile specimen has 7 thoracic chaetigers on the left (Fig. 49E), and 6 on the right (Fig. 49F). Collar chaetae of juvenile lack a boss; consist of shaft and blade (Fig. 49G). Abdominal chaetae with long shafts present throughout abdomen (Fig.  49E,F); their flattened, trumpet-shaped, distal ends have one side drawn out into a tapering process.
Remarks. The total lengths of Ishaq and Mustaquim's specimens (1996: 174) from Manora, Karachi Coast, Pakistan, ranged from 5.0-25.0 mm. For a discussion on the synonymy of Pomatoleios and Spirobranchus see earlier discussion in the present paper. As regards special collar chaetae their reduction is more than in the preceding species to to the extent of their lacking even a trace of a boss, consisting merely of a shaft and blade (Fig. 49G), and being indistinguishable from the bladed notochaetae of the remaining thoracic chaetigers.
Tube. Missing in all specimens. Mohammad (1971: 300) does not describe them, and provides only a note on occurrence of the in Kuwait and the world distribution.
Worm. An easily removable epizoic bryozoan was found on the bases of opercular processes of the three specimens (Fig. 51A). Colour of opercular processes translucent white, their spines transparent white. Operculum on left side, no rudimentary operculum; opercular plate circular, slightly dorsoventrally oval; bears three main processes, each with two dichotomies in 3 specimens (Fig. 51B-D); in one specimen, however, dorsolateral spines are branched thrice. A single medial spine is located at or below the second dichotomy, followed distally by one, two, or three along each branch (Fig. 51C). Peduncular wings fringed, with a double row of papillae (Fig. 51A,C,E). Radioles transparent and colourless, conspicuously transversely wrinkled (Fig. 51B,D), end in pinnule-free tips, 1 ⁄3 to 1½ times as long as longest pinnules. Maximum number of radioles observed on each side 23, arranged in a circle; inter-radiolar membranes fringed (Figs 51B,D). Thorax consists of 7 chaetigers; apron present.
Chaetae. Older bayonet-shaped special collar chaetae may possess abraded blades (Fig. 51F,G), whereas the blades of those newly formed within the collar fascicles are not abraded (Fig. 52A). In the latter, the boss is squarish; a long part of the shaft preceding the boss is serrated, nearly equal to the length of the blade and the blades end in tapering tips; an unserrated notch is absent. Thoracic uncini bear 8-12 teeth in a single row, and the anterior gouged process; abdominal tori lateral (Fig. 51B) anterior abdominal uncini possess 7-9 teeth and an anterior gouged process. Abdominal neurochaetae possess flattened trumpet-shaped tips, one side drawn out into a tapered process, and with long shafts projecting from their abdomen are present from the 2nd or 3rd anterior abdominal segment onwards (Fig. 51B).
Remarks. Spirobranchus sp. 4 superficially similar to S. tetraceros with regard to the branched opercular spines borne on a circular opercular plate. However, although the peduncular wings in S. sp. 4 and S. tetraceros are fringed, Figure 52. (A), Spirobranchus sp. 4: bayonet-shaped special collar chaetae from within collar fascicle. B-I, from type material of Spirobranchus maldivensis Pixell, 1913, BMNH 1924  those of the former possess a double row of numerous papillae (Fig. 50A,C,E), while those of the latter possess a single row of papillae (Fig. 46D,L). There also appears to be a difference in the maximum number of branchial radioles attainable, about 23 in S. sp. 3, but more, about 28-30 in the Kimberleys S. tetraceros, for instance. Further studies are needed S. sp. 3, including its tube.

Description
Measurements. From one of Pixell's types, as follows: total length 19.5 mm; length of operculum and peduncle 6.0 mm; long diameter of operculum 2.0 mm. Operculum plus 27 radioles on left side, no rudimentary operculum on right; 28 radioles on right side. Length of abdomen: 11.0 mm; number of segments 52.
Tube. In type specimens, caramel to faintly pinkish in colour, with slightly darker pink along median longitudinal ridge; latter prominent, with conspicuous forwardly directed projections which may be triangular towards the middle of tube, or end in laterally flattened tips with their stumpy ends appearing as if they were broken off. They are low anteriorly, and the median longitudinal ridge itself may be somewhat sinuous. All three types present on one of the tubes (Fig. 52B). Tube colour of specimens in additional material examined, BMNH 1939.9.2.554-556, range from light caramel to brownish and pink. Some have a pair of light brown longitudinal bands along each flank and a pinkish colour along the median longitudinal ridge. Otherwise, the form of the tubes and other characters agree with those of the type specimens.
Worm. Operculum in type material on left side; no rudimentary operculum on right; opercular plate circular, flat, lacks processes but shows concentric circular markings (Fig.  52E); peduncular wings broad and unfringed; 27 radioles on the left, 28 on the right in one specimen, arranged in a circle on each side; inter-radiolar membranes unfringed. One or two whitish patches present at the base of each radiole; uncertain whether they represent eyespots or mere pigment patches in the preserved material. Thorax: slender, its uncinal tori not borne on conspicuous pads (Fig. 52C,D,F).
Chaetae. Bayonet collar chaetae (Fig. 52G) differ from those of other species of Spirobranchus hitherto described in this account in being slender, their serrations comparatively inconspicuous along the blade, the boss and the serrated part of the shaft preceding the boss. Thoracic uncini: sawshaped (Fig. 52H), with about 12-13 teeth, in addition to the anterior gouge; anterior and posterior abdominal uncini similar, the former (Figs 52J) with about 10-11 teeth, and the latter about 10-12.
Remarks. Spirobranchus maldivensis can be recognized from most other species of Spirobranchus by its simple circular opercular disc, which is devoid of processes, the arrangement of its radioles in a circle, unfringed peduncular wings and inter-radiolar membranes, the form of its bayonet collar chaetae, and the caramel to pinkish colour of its tube.

Description
Measurements. in the above three specimens as follows. Total lengths 7.1-11.5 mm; thoracic widths, 1.5-1. Chaetae. Special collar chaetae bayonet shaped, with a prominent boss (Fig. 53F,G), sometimes with a constricted notch between boss and blade (Fig. 53G). As also seen in the latter figure, length of the blade is about 1½ times that of serrated distal part of the shaft plus boss. Thoracic uncini bear 13 or 14 teeth, anterior abdominal uncini 10 or 11, in addition to the anterior gouged process.
Remarks. Opercular plates of all three specimens had an epizoic hydrozoan coelenterate belonging to the genus Teissiera, Family Teissieridae on them (Fig. 53A), even completely covering the gaps between the opercular plates. A merely cursory examination could possibly result in the present material being identified as belonging to a species of Spirobranchus possessing a simple opercular plate, such as S. maldivensis, as indeed is evident from the original identification label on the jar. Removal of the epizoic coelenterate, however, reveals that it is not so. One of the three specimens had subsequently been cleared of the coelenterate which, according to the second label, led to the specimens being identified as Spirobranchus latiscapus. Clearing the remaining two worms during the present study revealed that they too belong to the same S. latiscapus-like species. Spirobranchus murrayi n.sp., differs from other "S. latiscapus-like" species having stacked opercular plates described in this paper with regard to the following characters. Firstly, with regard to their special collar chaetae: in S. murrayi they are bayonet-shaped, a boss is present, and the blade is longer than the serrated distal part of the shaft plus boss (Fig. 53G). In S. zibrowii they are not bayonetshaped, a boss is not distinguishable from the blade, and the distal part of chaetal shaft is not serrated (Fig. 63K). In S. zelandicus they are more or less bayonet-shaped, the boss is elongated and merges gradually with the blade, and the blade is about as long as the serrated distal part of the shaft plus boss (Fig. 61F); those of S. tenhovi (Fig. 61) are smaller and more slender than those of S. murrayi. As in the latter, the blade is longer than the serrated distal part of the shaft in S. tenhovi, but unlike in the former, the boss gradually merges with the blade and is barely recognizable.
Total length of operculum and peduncle is approximately equal to that of its abdomen in S. murrayi (Fig. 53B). On the other hand, the abdomen is significantly longer in other S. latiscapus-like species: in S. zelandicus and S. tenhovi: about twice as long as the operculum plus peduncle.
Furthermore, in S. murrayi, the ratio of the diameter of the basal opercular plate to length of opercular peduncle (based on those described in Fig. 53B,D), is approximately 1:2. In S. zibrowii (based those in Fig. 63J,G,H) approximately 1:1. In S. zelandicus (based on those in Fig. 62B,D,E) it is 1:1.3. In S. tenhovi, new species, (based on those in Fig.  63C-E) is 1:2.7-3.6.
Based on a comparison of the number of radioles in relation to total length of worm, S. murrayi is apparently also a smaller species. The largest of the three specimens is only 11.5 mm long, and its abdomen is 4.7 mm, comparatively broad (Fig. 53A), and has only 53 segments, but possesses 27 radioles on the left side and 29 on the right; number of its radioles exceeds that, for example, a specimen of S. zibrowii, which is over twice its total length (26.8 mm), has an abdomen with 88 segments, but only 21 radioles on the left and 22 on the right. In the new species S. tenhovi, to be described later, a worm with a total length of 34.5 mm and having about 75 abdominal segments has 33 radioles on the left and 33 on the right.

Spirobranchus paumotanus
Material examined. BMNH 1970.855, numerous specimens, including a tiny juvenile, 5; Solomon Islands, separated in Royal Society Expedition to the Solomon Islands, coll. and presented by P. Gibbs. Identified as S. giganteus (Pallas, 1766), according to label on the jar.

Description
Measurements. Maximum width of tube figured (Fig. 54A) Tube. White; mutually bonded and aggregated into clumps, their shape quite characteristic, when unhindered in their formation by competing individuals in the same clump. Each tube has an obtusely triangular cross-sectional appearance, and initially only a low median longitudinal ridge (Fig. 54A), then abruptly develops 1 to about 4 or 5 massive forwardly directed teeth (the one figured has two); most anterior tooth projects conspicuously over the aperture. A very wide lateral flange present on each side. Tooth-like processes at the anterior ends of tubes may be hindered from developing or may be obliterated by tubes of competing individuals growing over them. Apparently, once a competing individual grows over the aperture of another tube, the latter continues to grow underneath the former without developing any more such processes. Consequently, the tube growing underneath may only bear 1 to a few such processes posterior to where the competing tube grows over it. Shapes of such processes and the direction in which they point are also variable. The most anterior process, for instance, may be pointed, somewhat flattened, or shaped like a cobra's hood, and directed forward or laterally. Tube surface frequently pitted with irregularly shaped foramina (Figs 54A, 55G), evidently caused by a boring sponge; some tubes lack them.
Worm. The following description is based on examination of 8 adult specimens. Operculum on the left, no rudimentary operculum on right; opercula of 4 above specimens damaged; hence description based remaining 3, while a partly damaged operculum is also described from another. Opercular plate in all specimens circular, somewhat con cave, and bears three horns which arise from a short stout stem (Figs 54B-F, 55A-D). Two of the horns are dorsolateral, the third ventral. As seen in the figures, the dorsolateral horns are quite horn-like and curved outwards and ventrally. Each dorsolateral horn characteristically bears along its lateral edge, what is apparently a groove, but is actually covered by a brownish chitinous layer; it may sometimes also bear a small knob dorsomedially at its base, (Fig. 54B-D). The ventral horn is also characteristic. Viewed laterally, it is the tallest of the three, projecting conspicuously anteroventrally, then curved anteriorly, and ending distally in two main outwardly curved branches (Figs 54B,F, 55B). Each of the latter may, in turn, be dichotomously branched once or twice towards its tip to give rise to short processes, which are outwardly curved (Figs 54B-F, 55A-D). An operculum with the tip of the left dorsolateral horn damaged is shown in Fig. 54D,E.
Peduncular wings: unfringed (Figs 54B,C,E,F, 55A-D). Radioles numerous, arranged spirally on each side; number of whorls in each spiral mostly 4 on either side; 5 on either side in one specimen. Prostomial ocelli are not discernible. On the ventral side of the thorax, distances between corresponding pairs of uncinal tori decrease posteriorly leaving an inverted triangular area between them; the last pair of tori meeting mid-ventrally in worms fixed within their tubes; not seen in juvenile (Fig. 55H,I). An apron is present. Maximum number of abdominal segments counted, 91. Anal segment bears a pair of small rounded swellings in juvenile (Fig. 55K), lacking in adults.
Chaetae. Special collar chaetae bayonet-shaped, boss not squarish in side view, curved smoothly from shaft to the base of the blade (Fig. 55E); blade is considerably narrower than shaft; approximately equal in length to that of serrated distal end of chaetal shaft. Remaining collar chaetae: similar to special collar chaetae, but considerably narrower. Thoracic uncini: saw-shaped with 12-14 teeth in a row; anterior abdominal uncini similar, with about 12 teeth in a row; both in addition to the anterior gouge. Abdominal neurochaetae with long shafts projecting from the body wall are present from the third segment onwards. Indeed, they are quite long in the anterior segments, the length of the shaft alone being about 15 times as long as that of the distal trumpet-shaped part and its attenuated process (Fig. 55F).
Juvenile specimen. With a width of only 1.5 mm at its anterior end, the juvenile tube is similar to the posterior part of adult tube. It is obtusely triangular in crosssection, has only a smooth median longitudinal ridge, and anteriorly directed transverse furrows, those at the anterior end foreshadowing the tooth-like processes that have yet to be developed (Fig. 55G). A wide flange present along each side. Irregular foramina formed by a boring sponge present even at this young stage. Total length of worm 7.0 mm; operculum on left side, its two dorsolateral horns have the typical shape, but the ventral horn, which has yet to be developed, is represented only by a small knob (Fig. 55H-J). Opercular peduncle unfringed (Fig. 55H); radioles number 5 on each side, arranged in a circle, which is evidently the plesiomorphous condition. Length of thorax 3.5 mm; apron present; length of abdomen 3.5 mm, about half achaetous anteriorly, the remaining half with about 23 chaetigers; anal segment bears a pair of small rounded knobs (Fig. 55K), which are lacking in adults.
Ten Hove (1970: 14) suggested that two subspecies could be distinguished under the genus Spirobranchus, namely, S. giganteus giganteus (Pallas, 1766) from the western Atlantic and S. giganteus corniculatus (Grube, 1862), new combination, from the Indo-West Pacific. Following further investigations, ten Hove (1994: 112-113) referred the S. giganteus corniculatus-complex to the S. corniculatuscomplex, noting that the opercular plate is egg-shaped in the latter. As seen from the description of S. corniculatus (Grube, 1862) earlier in this account (Figs 38A-D, 39A-D,I,J), the opercular plate is indeed conspicuously wider anteriorly than posteriorly, and its outline is more or less like a longitudinal section of an egg or a pear.
In contrast, that of S. paumotanus is circular (Figs 54B-F, 55A-D,H,J), as it is in species such as S. tetraceros (Schmarda, 1861), (vide ten Hove, 1970: 6, and Figs 44C and 45G-I), as also in S. gardineri Pixell, 1913 and other species of Spirobranchus described in this paper. There are, however, also other differences between S. paumotanus and S. corniculatus with regard to the operculum. Each dorsolateral horn in the former is unbranched, anteroventrally curved, pointed, and has a characteristic posterolateral groove covered by a brownish chitinous layer (Figs 54B-E, 55A-D). In contrast, each dorsolateral horn in S. corniculatus usually ends in a pair of short pointed process, while also bearing a group of short processes dorsally and a short dorsal curved spine beyond the latter and lack the characteristic groove covered by a brownish chitinous layer (Figs 39D,I-J; 40A-C). The ventral horn in S. paumotanus is also characteristic, being the highest of the three and projects conspicuously, anteroventrally, and then curves anterodorsally. It ends in two main outwardly curved branches, which may, in turn, be dichotomously branched at its tip once or twice to give rise to short outwardly curved processes ( Fig. 54B-F). In contrast, a ventromedian process is frequently absent in S. corniculatus (Figs 39A,B,D,I,J; 40C), sometimes being represented by a knob (Fig. 39C) or by an aberrant process (Fig. 40A,B).
There are also differences with regard to tube colour and form, maximum sizes attained, maximum numbers of radiolar whorls, and number of teeth in the thoracic uncini. The tube of S. paumotanus is white and initially bears only a low median longitudinal ridge, which abruptly develops one to few massive forwardly directed teeth, with the most anterior one often projecting conspicuously over the aperture. On the other hand, that of S. corniculatus ranges in colour from yellowish orange to bright orange and has a single, somewhat smooth, median longitudinal ridge, with fine transverse ridges on either side of it. It does not bear similar massive teeth, with one of them projecting over the aperture of its tube. The maximum total length attained by the worm in S. paumotanus is about 35.0 mm, with a thoracic width of 2.5 mm, as opposed to over 110.0 mm and 10.0 mm, respectively, in S. corniculatus. A specimen of S. corniculatus from Malé described later in this paper has a total length of 90.3 mm and 226 abdominal segments. On the other hand, the maximum number of abdominal segments in S. paumotanus is probably less than 100. The maximum number of radiolar spirals observed in S. paumotanus is 5 on each side. In comparison, four out of the five specimens of S. corniculatus from Kimberleys collection AM W202960, possess 7 spirals on each side, and the fifth 5. Again, 10 radiolar spirals were counted in the very large specimen from Kimberleys collection AM W202958. There is also a significant difference with regard to the maximum number of teeth in the thoracic uncini. It is 12-14 in addition to the gouged process in S. paumotanus, whereas it is double that, 22-24, in S. corniculatus. Mörch (1863: 411, fig. 26) describes Pomatoceros bucephalus from Zamboanga, the Philippines, which is very similar to S. corniculatus (Grube, 1862). It is synonymized with the latter here, which has priority over the former. Gibbs (1969: 203) describes the habitat of S. giganteus as "embedded in Porites boulders (Gibbs, 1969: fig. 138) and encrusting on reef platform." However, information concerning the habitat of the Komimbo Bay sample, BMNH 1970.855 (= S. paumotanus [Chamberlin, 1919] in the present account) is quite different and indeed quite interesting, since it is unusual for a Spirobranchus species. According to the label in the jar containing the above specimens collected on 19/08/65, the species lives "Under boulders in muddy silt." The label also states "MTL-LWM, silt." Spirobranchus pseudopolytremus n.sp.

Material examined.
HoloTyPe and ParaTyPe, BMNH 1959.10.10.39; 2 specimens with only anterior parts of their tubes, one worm complete, the other incomplete posteriorly, Cronulla, New South Wales, 34°04'S 151°09'E, identified by D. Straughan as Temporaria polytrema (Philippi, 1844), and presented by B. Dew. Worms in this collection not too well preserved, apparently having been dried up previously and re-introduced into alcohol.

Description
Measurements. Maximum external diameter of tube 5.2 mm, including the lateral flanges; 3.6 mm, excluding flanges. Maximum height, including MLR, 5.0 mm. Total length of complete worm (holotype), 10.0 mm; width of thorax 1.5 mm; length of operculum and peduncle 3.0 mm; diameter of opercular cap 1.7 mm; and length of abdomen 4.3 mm. Branchial crown with 15 radioles plus operculum on left and 14 radioles on the right. In the second specimen (paratype) branchial crown has 18 radioles on each side, in addition to the operculum on the left.
Tube. Wide anteriorly with a broad flange, abruptly narrows posteriorly (Fig. 56A,B); MLR low smooth, somewhat irregularly formed, ends anteriorly as a small low tooth over aperture. A row of irregularly formed foramina present on each side of MLR, at its base (Fig. 56B). An irregularly formed, low, narrow, LLR present on either side of the MLR. A row of irregularly formed foramina present on each side at the base of each LLR. Faint, smooth, anterolaterally directed ridges and grooves, and foramina, extend into sides of tube and lateral flanges (Fig. 56A,B). Worm. Operculum on left side, no rudimentary operculum on right. Number of radioles: 15 on the left and 14 on right in holotype; 18 on each side in paratype. Operculum conical, its peak dorsally curved (Fig. 56C-G); although the peak is dorsally directed in holotype, it is a short cylindrical stump in paratype (Fig. 56C-E). A fine granular overlay can be seen at higher magnifications under certain angles of illumination (Fig. 56G). Opercular peduncle fringed (Fig. 56G). Dorsolateral and ventral views of the holotype presented in Fig. 56C,D. Radioles arranged in a circle; their pinnule-free tips thicker than and somewhat longer than preceding pinnules. Inter-radiolar membranes present; unfringed (Fig. 56H). Collar conspicuously broad (Fig.  56C,D); thorax consists of 7 chaetigers; prostomial ocelli and thoracic glands not discernible; apron, present (Fig. 56D).
Chaetae. Special collar chaetae stout and truncated, like acicula, serrated (Fig. 56 I); no indication of a boss or serrated part of shaft preceding it; at the commencement of serrated blade-like part, width of clear unserrated part of shaft is 2½ to 3 times that of serrated part of blade (Fig. 56 I). Chaetae accompanying special collar chaetae highly reduced, in comparison. Uncini saw-shaped; thoracic uncini with about 9 or 10 teeth, and posterior abdominal uncini about 11, arranged in a single row, in addition to the anterior gouged process; sides of latter pointed and directed outwards.
Remarks. According to a note in the jar by ten Hove, "taxonomic change to be processed = Pomatostegus polytrema." The latter was transferred to the genus Spirobranchus by Bianchi (1981: 107). As discussed in the remarks under Spirobranchus sp. 1, cf. polytremus (Philippi, 1844) earlier in this paper, it appears that there are a number of Spirobranchus polytremus-like species in the Indo-West Pacific area, which need to be sorted out. Imajima (1977: 102-106) remarks that "It is unlikely that" the species described by him "will prove to be the same as the Mediterranean species". He decided not to name his species in view of the complexity of the group involved, the confusion of names by previous authors, and the scarcity of data on infraspecific variation of the individual, and expressed the need for "a full-scale revision of the group." For the same reason, the S. polytremus-like material from the Kimberleys was not identified above. However, as seen from Fig. 40D-F in this account, and, for example, Bianchi (1981: fig. 41c,d) that, although the operculum of the Kimberleys and Mediterranean species are somewhat similar, there are important differences with regard to their tubes. The tube of S. pseudopolytremus n.sp., (Fig.  56A,B) is quite different from that of S. polytrema from the Mediterranean, e.g., as figured by Bianchi (1981: fig. 41a,b); while that of the former is wide anteriorly, and abruptly narrows posteriorly, it is long and sinuous in the latter.
The tube of S. pseudopolytremus is superficially similar those of two other new species are dealt with in the present paper, namely, S. elatensis and S. zibrowii. However, whilst the MLR in S. pseudopolytremus is low and ends in an inconspicuous process over the aperture (Fig. 56A,B), it is high, thick, and ends in a conspicuous projection over the aperture in S. elatensis (Fig. 50B) and S. zibrowii (Fig.  63A,B,D). The three species also differ amongst themselves with regard to their opercula. In both S. pseudopolytremus the operculum is conical, but whereas that of the former lacks any processes at its apex ( Fig. 56E-G), that of the latter bears a group of reduced S. tetraceros-like processes at its apex (Fig. 50C,D). On the other hand, the operculum of S. zibrowii bears 2 to a few super-imposed circular discs (Fig. 63E-J).
Although there are some similarities with regard to special collar chaetae among Spirobranchus sp. 3, S. zibrowii and S. pseudopolytremus, there are also differences. Those of S. sp. 3 lack a boss at the end of the chaetal shaft. However, the transition between the end of the shaft and the beginning of the blade can be recognized at the bend along the serrated side of the chaeta (Fig. 49D). As seen in the latter figure, another important character is that the serrated distal part of the chaetal shaft is longer than the blade. The special collar chaetae of S. zibrowii (Fig. 63K) differ from those of S. sp. 3 in that the distal part of its chaetal shaft is unserrated, and that it lacks a boss. Although the special collar chaetae of S. zibrowii and S. pseudopolytremus are superficially similar in lacking a serrated distal part of the blade, an elongated boss with a faint indentation between the latter and a reduced blade can be recognized in those of S. zibrowii (Fig. 63K) unlike in S. pseudopolytremus (Fig. 56 I). Moreover, those of S. zibrowii (Fig. 63J) are long and slender compared those of S. pseudopolytremus (Fig. 56 I), which appear to be aciculate.

Description
Measurements. For above specimens identified as S. semperi Mörch, 1861, by Pillai (1960, as follows: tube at aperture, 11.2 mm wide and 5.2 mm high, including the median longitudinal ridge; total length of worm 30.5 mm; abdomen 18.0 mm. In those identified as and S. semperi Mörch, var. acroceros Willey, by Pillai (1960): height of tube at aperture, 7.0 mm, including median longitudinal ridge; total length of larger worm, 17.5 mm, its abdomen 12.0 mm; Tube. Colour, in above specimens identified as S. semperi Mörch, 1861, bright pink in patches on a white background; in those identified as S. semperi Mörch, var. acroceros Willey, pinkish orange at aperture, secondarily overlaid with a light bluish-grey colour, apparently by an encrusting organism. As regards tube shape, both, those identified as S. semperi Mörch, 1861, and S. semperi Mörch, var. acroceros Willey, possess only a single, prominent, thick, sinuous median longitudinal ridge, which projects somewhat over the aperture (Fig. 57A,B).
Worm. In both S. semperi Mörch, 1861, andS. semperi Mörch, 1861, var. acroceros Willey, operculum on left side, no rudimentary operculum on right. Opercular plate: circular; bears 4 stiff chitinous horns, two dorsolateral, and two ventral. Each dorsolateral horn bears only a single very short dorsomedial branch; short, curved, stiff chitinous spines occurring along its branches, simple medially, mostly bifid dorsolaterally (Fig. 57C,D). Peduncular wings: fringed. Radioles arranged in a circle, 26 on each side in former, 24 on each side in latter; inter-radiolar membranes present and unfringed in both. Chaetae. Special collar chaetae of S. semperi bayonetshaped (Fig. 57E): blade narrow and tapering, about half as wide as the distal part of the shaft, about 1½ times the length of serrated distal part of shaft, boss squarish. Thoracic uncini saw-shaped (Fig. 57F) and bear about 15 teeth and an anterior gouged process. Special collar chaetae of S. semperi Mörch, 1861, var. acroceros are similar (Pillai, 1960: fig.   7B), except that the chaetal shaft on the opposite side of the boss is more curved.
Remarks. Spirobranchus semperi can be distinguished from S. tetraceros as follows. In S. semperi, tube colour is bright pink to pinkish orange on a white background, has a more or less circular outline, and a high, prominent thick, wavy MLR which ends anteriorly as a short pointed tooth over the aperture (Fig. 57, A,B). Operculum in the larger specimen bears four separate horns, the dorsal pair being longer than the ventral pair. Each of the dorsal pair of horns is branched only once, one branch being dorsomedial and the other dorsolateral (Fig. 57C,D), the former being very short compared to the other. The two ventral horns are un-branched. Inter-radiolar membranes are unfringed. The second specimen is similar, although its two ventral opercular horns appear to arise from a common base. In the two specimens identified above as S. semperi var. acroceros are similar, but for the fact that its four horns, which comparatively shorter, arise from a conical opercular plate. Its inter-radiolar membranes are also unfringed, and Figure 59. A-H, Spirobranchus sp. 5. (A) an entire worm; opercular horns are damaged but rest of the body is intact; shows unfringed peduncular wings, inter-radiolar membranes unfringed or bearing a single papilla; abdominal neurochaetae with long shafts projecting beyond body wall and anal segment bearing a pair of papillae; (B) lateral view of the opercular plate and horns of another specimen; dichotomies apparently similar to that of S. kuwaitensis, as also is the spine preceding the second dichotomy, which is directed towards the centre of the opercular plate; (C) operculum and spines viewed from anterior end: opercular spines, dichotomies, and form and sizes of the lateral spines are quite different from those of Spirobranchus semperi Mörch, 1861, (Fig. 56C,D); shows the fissures along the main branches of the opercular spines; (D) operculum and spines, and radioles, viewed from the anterior end; shows fissures along the main branches, and the medial spines. The inter-radiolar membranes possess only a single insignificant papilla; (E) branchial radioles: showing inter-radiolar membranes bearing single papillae; (F) operculum of a juvenile worm viewed from below, showing the presence of a central hollow in the opercular plate, and the unfringed peduncular wing. G,H, bayonet-shaped special collar chaetae: (G) an abraded chaeta showing its squarish boss; (H) a chaeta showing partly abraded boss, but is otherwise similar to G; blade in both abruptly tapered and only slightly longer than the serrated distal end of chaetal shaft plus boss. its bayonet-shaped collar chaetae similar. In S. tetraceros, on the other hand, the tube is white to light caramel coloured, obtusely triangular in cross-section, has a very low MLR and fine transverse growth markings extending laterally from it (Fig. 44A). Its operculum bears three main horns, of which the dorsolateral pair is branched thrice and the ventral horn is branched twice; the two branches of its dorsolateral horns are roughly equal to each other in size, as also are those of the ventral horn. The spines along the branches are somewhat longer, lighter in colour, not as chitinous and stiff as those in S. semperi. In its bayonet-shaped special collar chaetae (Fig.  45C,D), the serrated distal part of the chaetal shaft is shorter than the blade, and the boss is smoothly sloping rather than squarish. Inter-radiolar membranes are fringed (Fig. 45B).
Considering the sizes of the above two lots of specimens, there is every indication that S. semperi var. acroceros Willey is a juvenile form of S. semperi. There appear to be no significant differences between Serpula quadricornis Grube (1878: 275, pl. XV, fig. 6)

Description
Measurements. Maximum external tube diameter, 2.0 mm; total lengths of worms, 9.7-8.6 mm; width of thorax 1.3-1.5 mm. Lengths of operculum and peduncle, 2.5-3.0 mm, opercular disc plus horns, 1.0-1.2 mm, and diameters of opercular plate, 1.5-1.6 mm. Abdominal lengths, 5.5-14.4 mm. Numbers of branchial radioles range from 12 per side to 17 on the left and 18 on the right. Tube. Colour white; grows on coral (Fig. 58A), complete tube of a specimen, with much of coral newly laid over it removed, shown in latter figure; a row of transversely elongated chambers present ventrolaterally (Fig. 58B). Viewed from above (Figs 58C,D), has a low, more or less smooth MLR, and a flattened area on either side of it, the edge of the latter forming a lateral ridge. Obliquely transverse, smooth, shallow ridges and grooves on either side of the MLR may appear to be arranged in a row.

Worm.
Operculum on left side, no rudimentary operculum on right; opercular plate circular, bearing three main horns; up to two main dichotomies may be present on each (Fig.  59B-D). Accessory spines at the end of and along the main branches: elongated, tapering, and mainly single. Fissures are present along the main branches, even extending into the dichotomies (Fig. 59C), apparently not having closed up as in other species. Peduncular wings unfringed (Fig.  59A,F). As seen in transparency through opercular peduncle of a post-juvenile specimen (Fig. 59F), a hollow is present ventrally, towards the centre of the opercular plate. Maximum number of 18 branchial radioles counted per side, arranged in a circle; their pinnule free tips being somewhat longer and thicker than the pinnules that precede them (Fig. 59E). Interradiolar membranes unfringed or bear a single papilla (Fig.  59A,D,E). Thorax consists of 7 chaetigers; apron present; abdominal uncinal tori lateral; two short pygidial papillae present (Fig. 59A).
Chaetae. Special collar chaetae bayonet-shaped, with a square boss in those that have not undergone abrasion (Fig.  59G), less squarish in partially abraded ones (Fig. 59H); their blades broad towards the boss, taper abruptly, and are only slightly longer than, or equal to length of serrated part of shaft plus boss. Uncini: saw-shaped; thoracic uncini with 12 or 13 teeth plus anterior gouge; anterior abdominal uncini with 8-10 teeth plus anterior gouge.
Remarks. Spirobranchus sp. 5 from from Sorong, Indonesia, is superficially similar to S. tetraceros from the Kimberleys, but there are important differences between them. Although the tubes in both are white, there are differences in their form. Tubes of adult S. tetraceros are obtusely triangular and bear only a single low, thin longitudinal ridge, the MLR. On the other hand, tubes of S. sp. 5 are less obtuse, more or less circular in cross-section, and bear not only a prominent and smooth MLR, but also a LLR at each dorsolateral angle of the trapezium. However, the opercular processes appear to be similar. The coral habit of S. sp. 5 (Fig. 58A), is similar to that described by Smith (unpublished thesis, 1985) for his S. tetraceros species B, from the GBR with the tubes following "sinuous paths between and along branches that are invested with coral skeleton." Further work is necessary to determine whether the species that grows on coral tips in Sorong, Indonesia, belongs to S. tetraceros (Schmarda)

Description
Measurements. Maximum diameter of tube 2.5 mm; total length of worm 27.5 mm; length of operculum and peduncle 6.5 mm, length of operculum and its horn 2.5 mm, length of horn 0.6 mm, width of thorax 1.5 mm, length of abdomen 16.0 mm, number of abdominal segments about 115.
Tube. White, porous along sides, apparently caused by a boring sponge; a single simple, very low, median longitudinal ridge present.
Worm. Operculum arises close to the base of left radiolar crown, rudimentary operculum absent. Opercular plate circular, bears a somewhat dorsally located column which, in turn, bears three bifid horns, two of them dorsolateral, and the third ventral ( Fig. 60A-D). Each dorsolateral horn has a pair of short pointed spines, located dorsolaterally, close to its tip, and another shorter spine, dorsally, almost at its tip ( Fig.  60A-D). Lower branch of bifid dorsolateral horns, in turn, ends in a bifid spine with their tips pointing away from the opercular plate. Opercular peduncle elongated. Ventral horn ends in two spines, with their tips pointing away from the opercular plate. Peduncular wings unfringed (Fig. 60A-C). Number of radioles about 40 on each side, spirally arranged, with two whorls in each spiral; inter-radiolar membranes unfringed. Thorax has 7 chaetigers; apron present. Abdomen consists of about 115 segments.
Chaetae. Bayonet-shaped collar chaetae with distal part of the shaft, boss and blade serrated; the blade itself is 1½ as long as the combined length of the serrated distal part of the shaft and the boss in fully formed bayonet chaetae (Fig. 60E,F). Thoracic uncini saw-shaped, with 15 teeth arranged in a single row at one end of the (first) torus, and 11 at its other end, in addition to the anterior gouged process. Abdominal neurochaetae with flattened trumpet-shaped ends and long shafts projecting conspicuously beyond abdominal wall occur from first segment. Mid-abdominal neuropodia bear 2 or 3 chaetae with trumpet-shaped distal ends and one side drawn out into a tapering spine (Fig. 60G). Abdominal uncinal tori lateral; anterior abdominal uncini bear about 9 or 10 teeth, and posterior abdominal uncini about 10 or 11 teeth in addition to anterior gouge.
Remarks. Spirobranchus sp. 6 superficially resembles S. richardsmithi, but closer examination revealed the following important differences. The tube in S. richardsmithi possesses patches of mauve colour on a white to creamish background, being darker in some specimens that in others. As shown in Fig. 41A,J, it has 3 to 5 low longitudinal ridges. On the other hand, the tube in S. sp. 6 is white and bears a single, low, median longitudinal ridge only. Of the three horns arising from the central opercular column in S. richardsmithi, the dorsolateral pair is not dichotomously branched (Figs 41B-E,K,M,N; 42B,C,E,G); at the angle between the ventral and each dorsolateral horn, there is a short spine with its tip directed towards the opercular plate. The tips of the bifid ventral horn are also curved towards the opercular plate. On the other hand, the opercular horns in S. sp. 6 are very different. Firstly, all its three horns are bifid (Fig. 60A-D). Secondly, the lowermost spine on each dorsolateral horn is curved away from the opercular plate. Thirdly, tips of the bifid ventral horn are curved away from opercular plate. There appears to be a difference. In S. richardsmithi a maximum of 4½ whorls occurs in the branchial lobe of a specimen of total length 42.4 mm; 3½ spirals in a specimen of total length 24.0 mm. In comparison, the condition in S. sp. 6, of total length 27.5 mm, is intermediate between the latter two, its 40 or so radioles on each side are arranged in only two spirals. The maximum number of teeth in thoracic uncini also appears to be different, being about 17 or 18 in S. richardsmithi and only about 15 in S. sp. 3, in addition to the anterior gouged process.
The above material appears to belong to a new species. However, further work on additional material would be needed in order to confirm whether the differences observed between the specimen from Tizard Bank and S. richardsmithi come within the scope of intraspecific variations of the latter or not.

Fig. 61B-H
Spirobranchus latiscapus. -Monro (1939: 152) (Marenzeller); easily mistaken to be a single species, but actually consists of two genera, as follows: 8 empty tubes belonging to a Serpula crenatalike species (Fig. 61A), and two specimens belonging to the genus Spirobranchus. One of the latter had the worm intact within its tube; in the other, the worm had been extracted previously from its tube, and was found lying loosely among the specimens. They were separated as holotype and paratype for the new species described below; and their respective registration numbers: NHM ANEA 2009.20., 2009.21.
Tube. Colour, yellowish-orange to orange and white; a light caramel to yellowish-brown longitudinal band present along base of MLR. Its shape is quite unlike that of any of the other known species of Spirobranchus (Fig. 61B); is 2½ times as high as it is wide at its anterior end, where the MLR is highest; latter sinuous, comparatively very high and consists of long conspicuous, fin-shaped forwardly directed processes of varying lengths (Fig. 61B). Two very low, thin lateral longitudinal ridges present on each side, one very close to base of MLR, and the other along its flank. Latter, well formed towards anterior part of tube, gradually becomes less so from its middle posteriorly (Fig.  61B). Along base of MLR, on either side of it, is a row of oblique, oval to elongated, forwardly directed foramina. Likewise, along either side of each LLR, at its base, is a row of similar foramina (Fig. 61B).
Worm. Operculum on left, no rudimentary operculum on right; conspicuously elongated ( Fig. 61C-E), compared to that of the other "S. latiscapus-like" species dealt with in this account. Consists of well-formed stacked opercular plates, holotype with 2 ( Fig. 61C,D); paratype with 4 ( Fig. 61E); each opercular plate has a thin rim along its edge. Peduncular wings unfringed (Fig. 61D,E). Number of radioles in holotype, 33 on each side in holotype, arranged in a circle; inter-radiolar membranes short, about ¼ as long as the radioles, unfringed (Fig. 61F); pinnule-free tips of radioles slightly thicker than pinnules. Thoracic chaetigers number 7. Uncinal tori progressively increase in length posteriorly, leaving a clear inverted triangular area on ventral side of thorax; at ventral end of each uncinal torus there is dark pigment spot (Fig. 61G); apron present ( Fig. 61G). Anterior abdominal tori short, located ventrolaterally. Long-shafted abdominal chaetae present throughout the abdomen, starting from about 2 segments from anterior end.
Chaetae. Bayonet collar chaetae (Fig. 61H) lack a squarish or pronounced boss; position normally occupied by boss hardly recognizable; gradually merges with blade. Distal part of chaetal shaft preceding boss: serrated (Fig. 61H), its length together with that of boss, about equal to that of blade. Uncini: saw-shaped; thoracic uncini with about 13 teeth, and anterior and posterior abdominal uncini about 11 and 13 teeth, respectively, in addition to the anterior gouge.
Remarks. Spirobranchus tenhovi n.sp. differs from S. latiscapus (Marenzeller, 1885) with regard to several characters, although the former was identified by Monro (1939: 152) as S. latiscapus (Marenzeller, 1885). Firstly, tube colour is yellowish-orange in S. tenhovi, and its shape is quite characteristic; its MLR consists of high wavy, finshaped sections, the most anterior one projecting forward conspicuously over the aperture. A row of oblique basal foramina is present on each side at the base of the MLR as well as at the base on either side of the two LLRs on each side. On the other hand, according to the original description of S. latiscapus by Marenzeller (1885), its tube colour is pinkish-red; it has longitudinal ridges bearing pointed processes or pointed lamellae, the MLR being the most prominent; sometimes only faint ridges present, frequently fine small spines or small pointed lamellae; sometimes only a single lateral ridge on each side. The special collar chaetae of S. tenhovi do not possess a prominent squarish boss, being hardly recognizable and merging smoothly between the shaft and the blade, unlike in S. latiscapus, where they possess a squarish boss. According to the literature, it had previously been thought that there is a single species of Spirobranchus having stacked opercular plates, namely, S. latiscapus (Marenzeller, 1885). However, the present study shows that there is a group of species having stacked opercula belonging to the genus. They include S. murrayi and S. tenhovi described in this paper, as well as two more species, namely, S. zelandicus n.sp., and S. zibrowii n.sp., as seen below.
Etymology. Dedicated to Dr Harry A. ten Hove of the Institute of Systematics and Population Biology, Zoology Museum, University of Amsterdam, the Netherlands. We owe much of our current knowledge on the systematics, phylogeny and biology of serpulids to his many contributions, both individually and in collaboration with many colleagues, including the present author, working on the group, as well as to sharing his unpublished data with his colleagues in the search.for answers to problems concerning the taxonomy and relationships of serpulimorph polychaetes. ), but actually containing two genera/species. (A) one of the 8 orange coloured tubes of a Serpula crenatalike species mistakenly included with the new species of Spirobranchus, which also has yellowish-orange tubes. B-H, from holotype and paratype of Spirobranchus tenhovi n.sp; (B) tube of holotype (from which the worm was extracted during the present study); MLR high and consists of forwardly directed fin-shaped sections, most anterior one projecting conspicuously forward beyond the aperture; a row of oblique, forwardly-directed foramina present on either side, at its base. Two, low, narrow, LLRs present on either side of MLR; each also possesses a row of foramina on either side, at its base; (C) dorsolateral view of anterior end of holotype, with a two-tiered operculum; (D) ventrolateral view of latter; (E) left ventrolateral view of paratype showing its four-tiered operculum; (F) part of radiolar crown showing unfringed inter-radiolar membranes; (G) ventral view of thorax of specimen figured in C, showing a dark pigment-spot at the ventral end of each uncinal torus; (H) special collar chaeta; its boss is not squarish but merges gradually between distal end of chaetal shaft and blade; blade is longer than serrated distal part of chaetal shaft plus boss.  (Marenzeller, 1885): by Monro, Discovery Report, 1936: 192, and

Description
Measurements. (Holotype) tube 9.7 mm high at aperture, including MLR; latter alone 3.0 mm high; maximum width at its base, 5.7 mm; and 5.5 mm at its sides. Total length of worm 26.9 mm, its thoracic width 2.6 mm; length of operculum and peduncle 8.0 mm; operculum only, 1.6 mm, and its diameter 3.5 mm; length of abdomen 12.8 mm; number of abdominal segments 93.
Tube. Tube surface of adult (holotype) and juvenile (paratype) covered over by a calcareous encrusting organism, with more or less transversely arranged apertures (Fig.  62A). Upon removal of encrustation, tube colour beneath, white, with a faint caramel colour along MLR and fainter colour along transverse ridges and flanks. Height of tube, greater than its width; MLR high, continuous, sinuous, partly divided into triangular fin-shaped sections, some with nearly pointed tips, and projects anteriorly over tube aperture as a thick, prominent, forwardly directed process (Fig. 62A). Ventrolateral parts of tube chambered in places where not attached to the substratum.
Worm. Operculum on left side; no rudimentary operculum on right. Opercular plate may either be tiered (4 plates) as in the holotype (Fig. 62B,C), or possess a single plate as in paratype (Fig. 62D,E). Viewed dorsally, opercular base and wings of juvenile broad and rounded laterally, and attached to the thorax by a comparatively narrow stalk (Fig. 62D,E). Peduncular wings unfringed (Fig. 62B,D). Radioles in holotype number 37 on the left and 36 on the right; apparently in one and a half spirals in their contracted condition, but actually in a circle. Inter-radiolar membranes unfringed. Collar: broad (Fig. 62B,D). Thorax has 7 chaetigers; apron present ( Fig.  62B). Abdominal chaetae possess long shafts protruding from abdomen commencing about two segments from the anterior end; abdominal uncinal tori lateral (Fig. 62B).
Chaetae. Special collar chaetae bayonet-shaped; a long part of shaft serrated distally; boss elongated, with curved outline; blade narrower than shaft, tapering, its serrations extending to its tip, its length shorter than combined length of serrated distal part of chaetal shaft and boss (Fig. 62F). Thoracic uncini bear 12 or 13 teeth, and anterior abdominal uncini 9-11, in addition to the anterior serrated gouge. Anterior gouge characteristic: its sides pointed and directed outwards, while part in between is serrated (Fig. 62 I,J). Abdominal chaetae like flattened trumpets distally, with one side drawn out into tapering process (Fig. 62G,H).
Remarks. Spirobranchus zelandicus n.sp., as shown in the present paper, is among a group of species belonging to the genus Spirobranchus that possess stacked opercula, although, according to several identifications in the literature, it was hitherto thought that there was only a single such species, namely, S. latiscapus (Marenzeller, 1885). Evidently, a number of species similar to S. latiscapus have been identified as such merely on the basis of their stacked opercular plates. Among them is S. zelandicus n.sp., from New Zealand. It is similar to S. zibrowii from the Kimberleys in possessing a many-tiered operculum (Fig.  63E-J). However, the two species differ in the following respects. The tube of S. zibrowii is white to faintly bluish, and faint to darker-grey along the flanks. It has a high, thick, wavy MLR, which projects conspicuously beyond the aperture (Fig. 63A-D), and on either side of the MLR there is a very low, LLR. At the base of the MLR, as well as of each LLR, one either side of them, is a row small obliquely elongated foramina. On the other hand, the tube of S. zelandicus is white to faintly caramel coloured; and although it has a high, thick, MLR which projects conspicuously over the aperture as in S. zibrowii, it lacks LLRs and the rows of foramina characteristic of the latter (Fig, 62A). The operculum in S. zibrowii bears 2 to about 4 irregularly or incompletely formed and eccentrically laid plates, which may be on a short cylindrical column (Fig. 63E-J), and its opercular peduncle is fringed. On the other hand, the opercular plates in S. zelandicus are more regularly formed, and its peduncular wings are unfringed (Fig. 62B-E). The special collar chaetae of S. zibrowii are not bayonet-shaped, the distal part of the chaetal shaft is unserrated and a boss is absent (Fig. 63K), whereas, those of S. zelandicus are bayonet-shaped, possess a long serrated area, about as long as the blade itself, and prominent boss (Fig. 62F). Both S. zibrowii and S. zelandicus are different from Spirobranchus latiscapus (Marenzeller, 1885) and other species described in this paper in several respects, including tube colour and form and chaetae.
Based on the original description of S. latiscapus by Marenzeller (1885: 218-219), its tube colour is pinkishred; slightly triangular in cross-section, and has a median longitudinal ridge and one to several lateral longitudinal ridges bearing pointed processes or lamellae. Median longitudinal ridge is the most prominent. Sometimes only faint ridges present, or only a single lateral longitudinal ridge on each side, frequently only fine small spines or small pointed lamellae. Although Marenzeller had several specimens, the only complete had a total length of 27.5 mm (21.0 mm without its radioles), its operculum and peduncle 6.5 mm. Operculum on left, the number of stacked opercular Figure 62. A-J, Spirobranchus zelandicus n.sp., BMNH 1936.2.8.3047-3048, Two specimens, from New Zealand, Discovery Expedition Stn. 941, collected on 20.08.1932. (A) tube of holotype, with tube of a juvenile, paratype, on it. Both are covered over by a thin calcareous encrusting organism having scattered foramina over its surface. A high median longitudinal ridge ends in a conspicuous stout tooth projecting forwards over the aperture in the adult (holotype); yet to be formed in the juvenile tube (paratype); (B) whole worm (holotype), showing its 4-tiered operculum, unfringed inter-radiolar membranes, wide collar and long-shafted abdominal neurochaetae; (C) another view of the 4-tiered operculum of the holotype; (D) whole worm (paratype) with its single-tiered operculum, un-fringed peduncular wings and inter-radiolar membranes; (E) operculum of the paratype, another view; (F) bayonet-shaped special collar chaeta showing its boss and the long serrated part of the chaetal shaft preceding the boss. The gradually tapering blade is shorter than the serrated distal part of the chaetal shaft plus boss; boss is not squarish, yet recognizable, smoothly rounded and considerably narrower than blade; (G,H) distal ends of abdominal neurochaetae; (I) thoracic uncini showing serrated gouge; (J) enlarged view of serrated gouge. plates 4, peduncular wings unfringed. Number of radioles, presumably on each side, 25, arranged in a circle. In another specimen, the operculum and peduncle was 8.5 mm long, its operculum alone, 2.5 mm, and it consisted of 6 stacked plates; their number in all his specimens ranged from 3-7. Thoracic chaetigers numbered 7. Bayonet collar chaetae described as having a squarish boss; and thoracic uncini as having a dozen teeth. Monro (1936: 192) identifies the above material from "Discovery" Station 941 off New Zealand, as S. latiscapus (Marenzeller, 1885). It is, however, described as a new species, S. zelandicus in the present account. Although Monro does not describe its tube, the following data are provided with regard to the worm. A larger specimen is about 23 mm long, its width 3.0 mm; operculum consists of 1 to 5 thin stacked calcareous plates; special collar chaetae described as "with a short wide striated fin-like process at the base of the narrow anterior blade" and the thoracic uncini as having about 12 teeth in addition to the basal gouge. Monro's description of the bayonet collar chaetae as having a fin-like process at the base of the blade appears to be due to a misinterpretation of the chaetal structure. As shown in the present study of several species of Spirobranchus, the special collar chaetae could range from being bayonet-shaped and possessing a squarish basal boss, to those in which the boss is reduced to various extents and even in being completely absent, but a fin-like process does not occur in Spirobranchus. The part of the bayonet chaeta described by Monro as being fin-like is, therefore, the boss at the distal end of the chaetal shaft. The descriptive term "fin-like" is used for that fin-shaped part of the bayonet collar chaeta that is separated from the rest of the blade in certain other genera as, for example, Filograna Oken, 1815 among Serpulidae, and Spirorbis Daudin, 1800 among Spirorbidae (vide Knight-Jones & Knight-Jones, 1977).
Spirobranchus zelandicus differs from S. latiscapus (Marenzeller) in the following respects. The tube in the former is white to cream coloured, has a high, thick, wavy MLR, which projects conspicuously over the aperture, but lacks LLRs. In contrast, that in S. latiscapus, as described by Marenzeller (1885), is pinkish-red and has not only a MLR, but usually also LLRs, all usually bearing pointed processes or lamellae. Special collar chaetae of S. zelandicus possess a smoothly elongated non-squarish boss, whereas that of S. latiscapus is squarish.
Etymology. Named after its type locality, New Zealand.

Description
Tube. In syntypes, colour white to faintly bluish, with faint to darker bluish-grey along flanks, much lighter or white along flanges. Triangular in cross-section with a wide flange on each side ( Fig. 63A-D); MLR high, continuous, sinuous, and projects forward quite conspicuously over aperture; a very low LLR present on either side of MLR. A row of obliquely elongated to oval foramina present on either side of MLR, at its base; likewise, a row of foramina present on either side of each LLR, at its base (Fig. 63A,B); moreover, another row of foramina, present along each flank of the tube. Variations occur, (Fig. 63C,D), where the LLR may be reduced in places and only foramina may be irregularly present. Shallow, anterolaterally-directed grooves alternate with low ridges along the lateral flanges (Fig. 63A,B); flanges usually obliquely chambered (Fig. 63C).
Worm. Syntypes poorly preserved, brittle, apparently having dried up and re-introduced into alcohol. Operculum is on left side, no rudimentary operculum on right; bears 2 to a few obliquely stacked plates; most distal plate may be considerably smaller in diameter than basal plate, and separated from rest by a short irregular cylindrical column ( Fig. 63E-H). In a third specimen two plates present, including basal plate, eccentrically attached (Fig. 63I,J); a translucent rim present on each terminal plate; a granular calcareous overlay seen when illuminated from certain angles (Fig. 63G-I). Peduncular wings apparently fringed in a specimen which shows 3 papillate processes on one side (Fig. 63J).
Chaetae. Bayonet collar chaetae do not possess a squarish boss; instead, it appears to be elongated, gradually sloping and merging with a slender tapering blade (Fig. 63K); in lateral view, about half-way from proximal end of blade, clear part of the latter considerably narrower than serrated area (Fig. 63K); distal part of chaetal shaft apparently unserrated; or, perhaps, indistinguishable from serrated blade; chaetae accompanying special collar chaetae of similar length but more slender (Fig. 63L). Thoracic, anterior and posterior abdominal uncini saw-shaped, bearing only about 8 or 9 teeth, besides anterior gouged process; two sides of latter pointed. Abdominal neurochaetae with long shafts protruding conspicuously from beyond body wall present along almost entire length of abdomen, commencing from about 2 or 3 segments from anterior end; their flattened  (Philippi) and presented by B. Dew. (A,B) two lateral views of a tube; (C,D) two dorsal views of another tube; (E,F) two views of operculum of a specimen; (G,H) two views of operculum of another specimen; (I,J) two views of operculum of a third specimen; (K) special collar chaetae: its shaft lacks a distal serrated area and its boss appears to be elongated and merges gradually with a reduced blade; (L) a "capillary" chaeta found together with the special collar chaetae; it is slender compared to the latter, but otherwise similar.
trumpet-shaped ends have one side drawn out into a tapering spine. Abdominal uncinal tori: short and lateral.
Remarks. It would be useful to include some notes on examination of additional material since the syntypes are not well preserved. AM W3578 consists of a large cluster of about 6.0 cm in diameter, with numerous intertwined mutually bonded tubes. They are mostly bluish-grey. Tube lengths in such clusters are difficult to determine. Ridges and foramina are as in those observed in the BMNH sample described above, but mostly lacking LLRs. while the associated foramina are very poorly developed or altogether missing along most of the tube. In one of the larger specimens, total length of worm is 26.8 mm; width of thorax 2.0 mm; length of operculum and peduncle 4.6 mm; length of operculum 2.6 mm and length of abdomen 19.2 mm.
In AM W3635, tubes mostly white, few with light bluish-grey along the flanks; maximum width of one of the largest tubes 7.0 mm, including the lateral flanges, 4.2 mm without the flanges. Operculum occurs on left side; no rudimentary operculum on right. Many opercula possess a single concave disc. Radiolar arrangement circular, with 21 left side; 22 on right. Number of abdominal segments: about 88. Chaetae similar to those described above; long-shafted flattened trumpet-shaped neurochaetae present throughout the abdomen, starting from about 2 or 3 segments from anterior end.
Although specimens having two or more eccentric discs are included under the present new species, those in AM W3635 mostly possess opercula consisting of a single concave opercular plate. They appear to be somewhat similar to that of Spirobranchus sp. 2 described earlier from Eden. Unfortunately, the single specimen of Spirobranchus sp. 3 lacks its tube. Further studies are necessary to determine whether it is similar to that of S. zibrowii.
Both H. Zibrowius and Harry ten Hove (pers. comm.) suggested that the present author describe the present species, since it was unlikely that they would deal with it owing to their ongoing research commitments.
Etymology. It is named after Dr H. Zibrowius, formerly of Station marine d'Endoume, Marseille, who, in his study of the above specimens in Natural History Museum, and in an unpublished manuscript, considered it a new species; also in appreciation of his numerous contributions to our knowledge on serpulimorph polychaetes.