On certain factors concerned in the production of eye colour in birds

ThE work of I-lurst, iNettleship, and others i~ this country, and Davenport in America, has been concerned chiefly with eye colour in mammals, more especially in the human subject. The facts which Hurst established as to the genetic importance of the presence oi' absence of pigment on the anterior ruff'ace of the iris in man, seem to be applicable in the main also to the avian iris. I am unable however to find that much work has been done on eye cotour in birds fl'om the genetic s~andpoint, and the observations here recorded must be regarded as an attempt to carry the analysis of this problem a stage further. I showed ~n 1912 (see Nc~.ture, Sept. 19, 1912) Chat in birds, not only the ciliary muscle, but the dilator and constrictor muscles of the iris are composed of fibres of the striated or voluntary kind, and that the movements of the pupil in birds are apparently subject to voluntary control. This~we should expect if the iris muscles in birds are innervated by medullated nerve fibres fl'om the cerebro-spinal system. This difference in the histological structure of the intrinsic muscles of the eye-ball in birds and other vertebrates (with the possible exception of the Reptilia) has an important bearing on the evolutionary methods by which the pigmentation of the Iris has been brought about in these different orders.

ThE work of I-lurst, iNettleship, and others i~ this country, and Davenport in America, has been concerned chiefly with eye colour in mammals, more especially in the human subject. The facts which Hurst established as to the genetic importance of the presence oi' absence of pigment on the anterior ruff'ace of the iris in man, seem to be applicable in the main also to the avian iris. I am unable however to find that much work has been done on eye cotour in birds fl'om the genetic s~andpoint, and the observations here recorded must be regarded as an attempt to carry the analysis of this problem a stage further.
I showed ~n 1912 (see Nc~.ture, Sept. 19, 1912) Chat in birds, not only the ciliary muscle, but the dilator and constrictor muscles of the iris are composed of fibres of the striated or voluntary kind, and that the movements of the pupil in birds are apparently subject to voluntary control. This~we should expect if the iris muscles in birds are innervated by medullated nerve fibres fl'om the cerebro-spinal system. This difference in the histological structure of the intrinsic muscles of the eye-ball in birds and other vertebrates (with the possible exception of the Reptilia) has an important bearing on the evolutionary methods by which the pigmentation of the Iris has been brought about in these different orders.

HISTOLOGICAL.
The "Bull" Eye in Birds.
By this term I mean the eye which owes its black or dark colour to ~he absence of pigment on the anterior surface of the iris. The delicacy and translucency of the iris tissues allow the posterior uveal pigment to shine through, and this gives rise to an appearance of blackness. The Bull eye in birds thus comes under Hurs~'s definition of the simplex eye. In the Bull.eye t~he chick character is in fact retMned during adult life in ~he bh'd just as, in the blue eye in the human subject, the simplex chmac~er common to most infl~nts (of white as opposed to negro descent) persists into adult life.
The eye of the white fm~tail pigeon provides a good example of the ': Bu.tt" or simplex avian eye. There is an entire absence of black, brown, or yellow anterior pigment while the connective tissue cells and the striated muscle fibres m'e sufficiently delicate in structure to allow t, he posterior hexagonal brown or black pigment to show through, and thus produce the black effect.
Carefilt examination of the "Butt" eye in the tiring bird with a tens in a good light wilt show a red appemance, like a ruddy glow, covering t, he periphmat zone of the iris. This appearance is due to a rich ptexus of capillary blood ~,essets with thin walls in this situation, and is not due t,) the presence of any pigment cells.
It is important to remember that the majority of black or very dark brown irides in birds owe their, dark cotour to the presence of anterior iris pigment, and are not "Bull " eyes at all. Even among pigeons the "Bull" or simplex eye occurs chiefly in wholly white birds, which are genetically speaking recessive whi~es, ~hough no~ of course albinos. Bateson and others have shown that in fowls white feather colour is of two kinds, Dominant white as in the White Leghorn, and l~ecessive white as in the white l%ose-co~nb Bantam, the Silky fowl, and an extracted white obtained by crossing certain white with coloured strMns. It is interesting to note that while (as we should expect) the colour of the iris in the Dominant White Leghorn is red or yellow (gravel eye) the iris colour in the ~ecessive white l%ose-comb Bantam is not "Bull" but red or yellow.
In the Silky fowl also the black eye is not a "Bnll" eye but owes its dark eolom" to black anterior iris pigment. There are however special points about the black eye of the Silky fowl t~o which I shall draw attention later.
In the S~ock Dove (CoIumbc~ oesc~s) a superficial observation of the eye of the adntt~ bird' would record the eye as "Butt" or simptex, i.e. deficient in anterior pigment. Microscopic examination of sections of the iris shows however that the black cotonr is due go the presence of numbers of branching cells packed with dark brown or black l~igment granules. These cells are not only present on f,he surface of the iris but atso among the muscle-fibres and deeper tissues in which there are also cells containing yellow pigment. Thus the black eye of the St~ock Dove is an anteriorly pigmented eye and differs from the normal type of eye colour in pigeons which is white, yellow, orange or xed, producing the so-called pearl, gravel 02" ruby eye.
In the Rock Pigeon (Goltt~ba livia), which is regarded as the ancestral form, the iris colour is yellow 02-orange.

The Pearl Eye.
Next in grade of pigmen~ to the "Bull" eye comes the so-called pearl eye in pigeons, and the "Daw" eye as seen in the ]~alay Fowl. Anterior iris pigment is absen~ in the pearl eye, but the iris tissues are not translucent as in the "Bull" eye. They are crowded with granules which are themselves colourless but prevent the passage of transmitted light, and when seen by reflected light give a grey white appearance to the surface of the iris.
h[icroscopic examination of the pearl iris shows that this opacity rio transmitted light is due to the presence of these colourless, but more or less opaque, granules, which appear as clear refl'actile spherules of moderate size. When these are densely aggregated the interference with the passage of light may be sufflc{ent to produce an opaque or black effect. The cells which"contain these g~'anules are scattered around the capillaries and amongst the muscle-fibres, and they do not seem to have such definite.outlines as the cells which contain the ordinary yellow or brown pigment. It is indeed an open question whet~er these refl'ac~ile bodies should be classifed as pigment granules, since under dark ground illumination they appear to be of a dull white colour.
Among Fowls the eye which most nearly corresponds to the pearl eye of pigeons is the so-called "Daw " eye of the Malay Fowl.
In the "Daw" eye the iris is free fl'om antra{or yellow or brown pigment, but ~he connective tissue ceils and the muscle fibres are loaded with granules which produce a tissue opacity and prevent the areal pigment fl'om shining through the iris, thus giving a grey appearance to the iris on reflected light. The pearl eye, and perhaps the daw eye, may in fact be regarded as representing the first or lowest grade of iris pigmentation, a condition of cell-opacity without cell-pigmen~, intermediate between the bull eye with its tz'anslucent tissue-cells and absence of pigment-granules on the one hand, and the fully pigmented eye with its connective vissue cells loaded wi~h yellow, brown or black pigment granules on the other.
The pearl eye in pigeons seems to con'espond in pigmentary gradation with the china-white or "W~lt" eye in dogs, horses and pigs. We must not lose sight of the fhct however that the substitution of s~ri~ted muscle-fibres in the avian iris fbr the plain muscle cells of the mammalian iris introduces a different histological factor in these different orders. In the mammatia there is every grade of blue eye fi'om the deep violet or almost black shade of the newly born child or kitten to the steel blue eye of the human adult, or the bluish white iris of the blue Dutch rabbit, or the china white iris of the "wall" eyed pig.
The paleness of the blue cotour or greyness of the iris depends on the granutm" condition or opacity of the tissue cells and the degree to which 'the pos[erior pigment is prevented f~om shining thr6ugh. Thus, an iris composed of translucent cells gives a butt or black eye, while an iris of which the cells are slightly opaque is blue, and an iris composed of wholly, opaque' granular cells is china white in cotour. All these irides are free from anterior pigment. The different reaction to incident light is'due to structural and not pigmentary differences, and tO the presence in the cells of these colourtess granules.
Genetically the pearl eye in pigeons acts as a recessive to the red or gravel eye. The daw eye in fbwIs is also a recessive to the amber eye, and to the black eye when the blackness is due to the presence of anterior pigment.
The Yellow or "Gravel" Eye.
In pigeons the yellow eye is.due to a ne~-work of branching cells crowded with small spherical yellow granules. These cells lie on the anterior surface of ~he iris: they cover the capillary blood vessels and surround the striated muscle fibres which in the pigeon m'e themselves free from pigment. The difference between the grey-white or pearl eye and the yellow or gravel eye in pigeons is essentially a structural difference. In the pearl eye the granules which fill ~hese connective tissue cells are chalky white to reflected and opaque to transmitted light, Whereas in the yellow eye the granules appear yellow both when viewed by reflected and transmitted light. If however the ceils which contain them are closely aggregated and prevent the passage of light, then the granules appear black.
Thus the yellow eye represents the first grade in iris pigmentation. In fowts on the other hand the gravel or yellow eye presents a different problem. In some breeds the orange cotoui" is due to the presence of large numbers of branclling connective gissue ceils containing yellow or yetiow:broivn pigment granules, which: tie around the capillaries and muscle-fibres. In other breeds, of which the Dorking and Orping~on are examples, the muscle-fibre.celts themselves also contain the yellow pigment granules. Bearing on this point, namely the presence of pigment granules in striated muscle cells, it is interesting to note that in some orange eyed biMs the striated cells of the ciliary muscle also contain yellow pigment, and this gives in some c,uses a distinct yellow cotour to the muscle which is quite recognisabte to the naked eye. When we recall the fact that the ciliary muscle has the same embryological origin, and remains throughout individual development in close structural and fimctionat continui.ty with the muscles of the iris Which act on the pupil, it is a matter of interest to find that both share in the pigmentary changes concerned in the production of eye cotour in some birds. It is clear that the occurrence of pigment in the voluntary muscle fibres entirely under cover of the sclerotic and unexposed to the action of light cannot have any influence in sexual selection. It must be regarded like many other pigmentary phenomena as due to an overflow of normal metabolic activity into cells which under ordinary conditions are free fl'om such changes. The deposition of pigment granuies in the iris muscle and in the ciliary muscle is chiefly found in domesticated birds. I have not, so far, come across pigmentation of the ciliary muscle in any wild bird. Such pigmentation is in ,fact a metabolic abnormality since it occurs in cells which in wild birds are normally fl'ee from pigmentary deposition. The effect of the deposition of a mass of pigment granules on the functional activity of a striated muscle cell is also a point of interest. It would be interesting to know whether the change affects or/ly the older cells, it does not at any rate affect all the muscle cells even in a heavily pigmented iris. Bearing in mind that the yellow iris of the domesticated fowl is produced by the interaction of several factors: (1) the presence of yellow pigment granules in connective tissue cells, and' (2) the deposition of yellow pigment granules in striated muscle cells ; it becomes a point of' interest to ascertain whether the yellow iris in wild birds owes its colour to the same or different factors. The owls. afford perhaps the best examples of the purely yellow iris among wild birds.
In the Eagle Owl (Bubo bubo)~/he peculiar velvety bright yellow appearance of tile iris is produced by a heavy coating of the front ii~ce of the iris with numbers of roundish or owd cells crowded with bright yellow spherical pigment granules. These cells are present also in Scops Owl (Otus scops). Probably in most, if not all of the yellow" eyed owls, the yellow cotour is due to the presence of these cells. ~The cells themselves are fairly constant i_n size and outline. They are free from 5--5 branching processes and gradually toss their cotour when submitted to the action of 5 ~ formalin solution.
In the Brown Owl (St~'ix alues) and the Grey Eagle Owl, and some other brown eyed species, the brown cotour of the iris is due to the presence of branched pigment cells containing brown granules. The same occm:s in the Eagles (Aquilinae), the Kites (Milvus) and some other birds of prey. The Egyptian Kits (Milvus cegyptius) and Batstsur's Eagle (Helota~'st~s ecaudatus) both tla~s dark brown irides, and in both the pigment is contained in br~nch cells. In.no case at present has any deposition of yellow Or brown pigmeng granules been observed in the striated muscle-fibre cells in the iris of any wild bird.

The Brown and Black Eyes.
Amongst Birds, as amongst Mammals, quite a large number of species possess anterior iris pigment which passes through all grades of brown up to black. The black iris is associated (with some exceptions, notably the Silky Fowl) with black feather cotour, and in its deeper grades with black leg colour. Histologically the brown (in its darker shades) and the black iris are produced by a well-defined layer of characteristic branching cells, which contain d,~k brown or black pigment, on the anterior surface of the iris. These cells intercommunicate by their branches and form a plexus of pigmented cells thickest over the capillaries in the peripheral or middle zones of the iris.
The cells are much alike in different species. The body of the cell and its processes are usually crowded with pigment granules of a rounded outline and fairly uniform size. Under favourable conditions a central nucleus can be made outmore or less fi'ee from pigment. These pigment cells permeate the thickness of the Ms wall for some depth and are often found encircling the striated muscle cells. They are structurally continuous with the branching pigment cells in the sclerotic .and the ,outer surface of the choroid behind the cornea, and are quite distinct from 'the hexagonal cells on the posterior surface on the iris. As has been already stated, they are especially numerous and heavily pigmented in birds of dark or black plumage. They are found in nestlings belonging to the Corvidae and allied species. They are also present in the newly hatched chick in the black breeds of fowls.
Genetically the black eye with its deeply pigmented branching cells is dominant over the yellow ills and other grades of:iris pigmentation thus : The daw eyed Matay hen crossed with the Black Orpington cock gives F~ with black plumage (with some red in the males), and all have more or t~ss black anterior iris pigment due to the presence of these characteristic branching ceils. As the F~ chicks develop however interesting changes take place. Among the limited nmnber of bi,~ts reared up to thepresent the putters retain the btact'~ eye in adult life, while in the cockerels the iris gn-aduatty assumes a yellow colour and at the age of 9 months or eartier, if the birds are sexually mature, the black becomes a gravel or orange eye. Thus the black ilqs of the hen and the orange ,of the cock in this cross are sex timited characters as in the ease of other gattinaceous birds. In the Golden Pheasant (Chrysogophu, s pictu, s) the iris is brown or brownish black in the female and bright yellow in ~he Ntty developed mate, and to a much tess marked degree the same is ~rue of the Mongolian Pheasant (Phasia~zus mongolicu, s). An interesting problem arises .as to the way in which this change in iris cotour is brought about in the adult mate. The transition is one from an epistatic to a hypostatic charactel, that is from a higher to a lower grade of p{gmentation. The chicks of both sexes of this Malay cross have brown black irides due to the presence ~of ttiese anterior pigment-containing cells and they retain this character while sexually immature. The change to the yellow colour in the developing cockerels occurs in patches on the surface of the iris and seems to be due to the removal of the cells containing brown or black pigment and of the substi-;ntion in their place of cells containing yellow pigment granules together with (i n the ease of the half-bred Malay fowl) the deposition of yellow pigment gTanules in the striated muscle cells of the iris. The question arises as to the disappearance of these brown pigment cells. Is it the result of atrophy and absorption or of migration to deeper parts of the iris ? There are re~ons for thinking that both factors are conee~ned in fhe process. As the iris assumes ~he yellow colour the ceils which contain the black or brown pigment eoincidently lose their dendritic processes and tend to become rounder in outline. Iu the Herring Gutt (Larus argen-~atus) the yettow iris is not fully developed in the adult mate until the fourth year. A careNt comparison of sections of the it;is in the nestling, ~he young mate, the young femate, and adutts of the two sexes of'the Herring Gull shows that gtie btaek cotour of the nestting's eye is due to the presence of a ptexus of branched cells containing brown or black pigment on the anterior surface of the iris. The female retNns more or tess of this dutt brown cotour during adult tife, but in the adutt mate a tayer of ceils containing yellow pigment replaces the layer of black pigment cells on the front of the iris, and these latter cells are found at a deeper level just above el-anterior to the uveal pigment cells. The appearances in the .Herring Gull suggest a downward migration of the black branched cells and a surfi~ce migration of the yellow ceils. This view that the coIour change is the result of ceil migration receives some confirmation in the case of young Rooks and other members of the Corvidae. The anterior surface of the iris in the nestling Rook (Gorvus ]~'ugilegus) is dotted with a number of branching ceils loaded w~h intensely black pigment granules. These cells are most numerous at the scleral margin of the it'is and are directly continuous.with a plexus of similar cells which lines the sclerotic on its choroidal surface. As the nes%ling, grows these cells become more numerous and spread over the anterior surface of the iris, as if m~grating towards the pupil from the scleral margin.

The "Black" Eye in Fowls.
If the Black Oi:pington (black eye) male be crossed with the Buff Orpington (gravel eye) female the F~ chickens all show some black anterior iris pigment on a yellowish background. In the pulleSs the brownish black colour persists in adult life while in the cockerels the black gives place to the yellow or gravel eye. Although in this cross the yellow eye seems to be a sex limited character the sharpness of the limitation is no~ absolute, for a few of the cockerels show in adult life a surihce layer of brownish pigment on a yellow background. In the Malay (daw eye) female crossed with the Black Orpington (black eye) male the cockerels also develop a yellow eye when sexually mature while that of the pullets remains black or dark brown, and in this cross the character seems strictly sex limited. In both crosses the deeper yellow pigmentation is due to tw~o factors as in the yellow eye in other breeds of fowls. These ikctors are (a) the presence of connecvive tissue cells containing yellow pigment granul~s and (b) the deposition of yellow pigment granules in the striated muscle fibres of the iris.

The "Triplex" Eye or the Eye containing two l~i~ds of anterior _[~is pigment.
Mention has already been made of Hurst's classification of the Duplex eye as one in which anterior iris pigment of some kind either yellow, brown, or black is present. In connection with this point I wish to draw attention to a group of eyes in Fowls of which the Silky, the Croad Langshan, the Houdan and some other breeds provide examples, which might be called Triplex eyes. There are, however, some objections to HuI~t's classification of eyes as Sir~plex and Duplex, and it might be better to divide eyes from the colour standpoint into (a) non-anteriorly pigmented and (b) anteriorly pigmented eyes, while (b) would be further divided into singly pigmented (Hurst's duplex) and doubly pigmented or (triplex) eyes. If the eye of the Silky Fowl be examined in the fresh condition, or weferably after a few days preliminary harden!ng in formalin, and if the anterior surface of the iris be exposed by careful peripheral detachment from the scleral" margin under wa%er and if further the antelior layer of black pigment cells be em'efullyscraped away with the point of a scapel from the front of the exposed iris, an underlying layer of redd!sh yellow pigment comes into view. This yello~y layer can in its turn be scraped away, leaving exposed the unpigmented stroma cells and the muscular fibre cells of the deeper portions of the iris, and if these be removed the posterior-uveal pigment is reached. The histology of this double layer of pigment cells of different colours in the Triplex eye is of interest.
The anterior black layer is made up of a dense network of ramifying. cells crowded with dark brown or black spherical granules, uniform in siz e , such as we have already described as occurring in the black eyes of many species of wild" birds. These cells are especially numerous arou/ad the capillaries on the anterior surface of the iris. The yellow or red-yellow layer lies beneath this and, as in the yellow eye of other breeds of fowls, is made up of two factors: (a) connective tissue cells containing large numbers of yellow pigment granules (these cells are less branched than the black ceils and are situated around the capillaries in striated muscle fibres); (b) a. copious deposit of yellow pigment granules in the striated muscle cells. There is a marked discontinuity and absence of grading between the yellow and black pigment layem. An object lesson is thus provided in the a~lult individual zygote of. the definite histological deposition of one unit characte5 viz. black pigment over a lower grade unit character, viz~ yellow pigment, and a realization in the individual ofthe factorial composition of the gametes into epistatic and hypos~atic wecursory factors.
Experiments have been carried out to test the behaviour of these two unit characters, black and yello w pigment, in the same eye by cross breeding. Five black-eyed Silk)" hens crossed with a gravel-eyed Old English Game Bantam cock gave in the F1 generation gravel-eyed cocks and mostly black over yellow-eyed hens. In the F2 generation some segregation occurred with a regrouping of the black and yellow characters, witli the result that both were found associated with white and game colou1" and with fluffy and close-feather pattern. The sex limitation also persists in the F~ generation, the black over yellow or t1{plex pattern being associated with ~he female and the yellow or duplex pattern with the male sex. In this case a secondary male sex character, yellow eye colour, is brought about by the toss of an epistatic character, black pigmen% uncovering as it were a hypostatic character, yellow pigment. The ultimate causes of the shedding of this epistat~c character must probably be sought for in the metabolic processes associa~.ed with the development of sex. The metabolism which develops yellow pigment g~'anutes ~nstead of black in connective tissue cells, and also brings about the deposition of yellow pigment g~'anutes in striated muscle cells, is no doubt associated with s functional acISivity of certain endocrinous glands among which we must ~nclude the primary sex glands.

The Triplex EyY or Eye with double anterior pigment in Wild Birds.
The black over yellow or triplex arrangement of eye cotour in the Silky Fowl is a normal example of epistatic character, black, overlying a hypostatic character, yeiiow pigment. Attention has already been directed to the fact that in the adult male Herring Gull the opposite condition is found. In riffs case a tayer of bright yellow cells cover~ the anterior surface of the iris and overlies a layer of black branched pigment cells. Thus yellow becomes epistatic to black in the adult mate bird. As we have already seen, this result is brought about pardy by the removal, by atrophy, of the black cells, and pardy by the migration of these cells to a deeper layer of the iris, thus uncovering the yetbw pigmeat cells. This arrangement also constitutes a sex limited character in other species. In the adult female EIerring Gull the brown colom" which is common to nesding birds of both sexes persists through life, while in the m~q]e when fully mature the iris assumes a yellow colour.
These changes in eye-co}our dependent on sex and age acquire additional interest from the fact that we have here a demonstration of an actual process going on in the tissues of the zygote during ontogeny, of which the hidden counter-part has previously taken place in the factorial constitution of the germ cell at an earlier stage.

The "Ruby" Eye in Birds.
We may first deal with the Ruby eye as it occurs in pigeons and doves, because the fi~ctnrs concerned in the producSion of this iris colour are 1lot tt~e same in nit species.
Ill the common Ring Dove (Tu~'tur com~z~uis) ~he ruby colou~" is due to the presence of a surface layer of branching cells with yellow pigment granules, which surround the capillaries on the anterior surface of the iris.
The rich ruby eye of the Victoria Crowned Pigeon (Goura victo~'iae) is due to ~he same cause. In this species also, as in the Doves, the red colour is common to both sexes. In the nestling Victoria Crowned Pigeon the colour of the iris is black. At the age of three mon~hs the iris assumes a bright yellow eolour, but I have not been able to ascertain the age at which the yellow gives place ~o the ruby colour.
In the Dragoon Pigeon and some other fancy breeds all grades of the ruby eye occur, from bright red through dull red to orange and yellow. In some birds the ruby or "red curran~" coloul-is limited to the outer zone of the iris where blood vessels abound, While the inner zone surrounding the pupil, where there are fewer capillaries, retains the orange or yellow colour.
The Ruby eye in Doves and Pigeons loses much of its brilliant red colour after death, and, as we shall see later when dealing with the chemical aspect of the problem, the red and eventually the yellow" colour are both lost when the eye is placed in a reducing agent like formalin.
The ruby colour is thus a superimposition effect due to the covering of red blood vessels with yellow pigment cells. The exact way in which the optical effect is produced is a subject for further study.
The Ruby eye in the Cayenne Lapwing (Vanellus eayenTzensis) presents a different problem. Here the red colour of the iris is due to the actual pi'esence of rounded, oval, or slightly branclied cells packed wi~h fine granules of a reddish mauve coloured pigment. These cells are situated on and among tile muscle cells and the capillaries of the iris. This is the only example at present found of a colour effect in the iris due to the presence of pigment granules which could not be included in tl~e yellow, brown or black groups. Not only are the granules which produce the red colour in the Cayenne Lapwing's iris histologically different fi'om the yellow pigment granules which form the surt~ace layer in the Dove's and Pigeon's iris, but they also differ in chemical composition. In the Lapwing the ruby colour persists after death and the pigment granules retain their red colour even after a long immersion in 5 ~ formalin.
In some of the Birds of Paradise, notably in Lawes' Bird of Paradise ~Parotia lawesi), and possibly in other species, striking colour effects are produced by a differential colouring of the iris. In Lawes' Bird an inner ring of deep iridescent, blue is sun'ounded by an outer ring of yellow pigment. This co]our effect is brought about in two ways. If a vertical section of the iris taken at right angles to the surface be examined under a low power the iris in the outer or yellow zone will be found to be four ol-five times thicker than in the inner or blue zone. This increased depth of tissue in the peripheral zone, containing as it does the muscular fibres and capillaries, partly accounts for the opacity of the structures and prevents the posterior pigment fi'om shining through, hence the absence of a blue effect. In addition the anterior suri~ce of the iris in this zone is covered with a layer of yellow pigment cells.
The inner or blue zone on the othm" hand is a mere thin membrane, sharply defined fi'om the outer zone by a step down as it were in thickness on the anterior surface of the iris. The thinness and translucency of the tissues in this inner zone readily allow the black uveaI pigment to shine through. The greater part of the blue effect is no doubt due to the presence of this underlying black pigment seen through translucent tissues. This however is not the whole explanation. A large portion of this tSin im~er zone consists of very finely fibrillated spindle cells of connective tissue type. The ends of these cells break up in hardened specimens into fibrillae which separate into wisp-like fringes. These elongated cells lie over the muscle fibre cells, and if the anterior surface of the iris in this situation be examined under a low power the parallel fibrillae bring about a diffraction grating effect. This accounts for some of the blue reflection of incident light. This can be shown by varyhlg the angle at which the light.is allowed to strike the surface of,the iris. At certain angles the blue effect is produced quite independentiy of the posterior black pigment and can indeed be obtahled afSer this has been removed. Thus, in the case of L,~wes' Bird of Paradise a brilliant colour effect is produced by a combination of three factors: (t) Thinness and translucency of iris tissue allgwing the uveal pigment to shine through.
(2) The absence of anterior yellow pigment cells in the inner zone.
(3) A peeulia!" physical conformation of the connective tissue cells in this area. k parallel fibrillation of cells acts as a diffusion grating and causes light to be reflected from the anterior surface of the iris at a certain angle as blue in colour It would be interesting to extend this enquiry to other species of Birds of Paradise and to individual birds of both sexes at different ages.

~HEMICAL.
I regret that I am unable to bring forward any detailed account of ~he difference between the black, brown, yellow, ruby, and pearl pigment granules from the chemical point of view. There is one definite and constant character which serves to distinguish the yellow froln the black granules and that is their different behaviour to reducing agents like formalin. The yellow pigment granules lose their colour after immersion in 5 ~ formalin for two or three weeks. A coalescence of the small granules into larger globules of yellow lipoid-like material takes place, and these larger globules may eventually merge into larger masses~ In some specimens needle crystaIs have formed inside these yellow oiI-Iike globules (see tIerring Gull). The black pigment granules which fill the branching cells that are so characteristic of the anterior layer in the black iris of many biMs are, on the other hand, highly resistant to the action of formalin; for 5 ~ and even stronger solutions have little or no effect in dissipating the colour. The granules do not tend to coalesce into larger globules nor do they show any signs of crystalloid change. The brown pigment granules are intermediate in their chemical reactions between the yellow and the black, and are partially affected by formalin.