Chapter XVI: Organs of Vision (2)

The lens when first formed is an oval vesicle with a small central cavity, the front and hind walls being of nearly equal thickness, and each consisting of a single layer of elongated columnar cells. In the subsequent stages the mode of growth of the hind wall is of precisely an opposite character to that of the front wall. The hind wall becomes much thicker, and tends to obliterate the central cavity by becoming convex on its front surface. At the same time its cells, still remaining as a single layer, become elongated and fibre-like. The front wall on the contrary becomes thinner and thinner and its cells flattened.

These modes of growth continue until, as shewn in fig. 289, the hind wall _l_ is in absolute contact with the front wall _el_, and the cavity thus becomes entirely obliterated. The cells of the hind wall have by this time become veritable fibres, which, when seen in section, appear to be arranged nearly parallel to the optic axis, their nuclei _nl_ being seen in a row along their middle. The front wall, somewhat thickened at either side where it becomes continuous with the hind wall, is now a single layer of flattened cells separating the hind wall of the lens, or as we may now say the lens itself, from the front limb of the lens-capsule; of the latter it becomes the epithelium.

The subsequent changes undergone consist chiefly in the continued elongation and multiplication of the lens-fibres, with the partial disappearance of their nuclei.

During their multiplication they become arranged in the manner characteristic of the adult lens of the various forms. The lens-capsule, as was originally stated by Kölliker, appears to be formed as a cuticular membrane deposited by the epithelial cells of the lens.

The views of Lieberkühn, Arnold, Löwe and others, according to which the lens-capsule is a mesoblastic structure, do not appear to be well founded. The contrary view, held by Kölliker, Kessler, etc., is supported mainly by the fact that at the time when the lens-capsule first appears there are no mesoblast cells to give rise to it. It should however be stated that W. Müller has actually found cellular elements in what he believes to be the lens-capsule of the Ammocoete lens. Considering the degraded character of the Ammocoete eye, evidence derived from its structure must be accepted with caution.

The vitreous humour. The vitreous humour is derived (except in Cyclostomata) from a vascular ingrowth, which differs considerably in different types, through the choroid slit. Its real nature is very much disputed. According to Kessler's view, it is of the nature of a fluid transudation, but the occasional presence in it of ordinary embryonic mesoblast cells, in addition to more numerous blood-corpuscles, gives it a claim to be regarded as intercellular substance. The number of cells in it is however at best extremely small and in many cases there is no trace of them. In Mammals there appear to be some mesoblast cells invaginated with the lens, which are not improbably employed in the formation of the vessels of the so-called membrana capsulo-pupillaris. In the Ammocoete the vitreous humour originates from a distinct mesoblastic ingrowth, though the cells which give rise to it subsequently disappear.

The development of the zonula of Zinn in Mammalia, which ought to throw some light on the nature of the vitreous humour, has not been fully investigated. According to Lieberkühn (No. 373, p. 43), this structure appears in half-grown embryos of the sheep and calf.

He says "At the point where the ciliary processes and the ciliary part of the retina are entirely removed, one sees in the meridian bundles of fine fibres, which correspond to the valleys between the ciliary processes and fill them; also between these bundles there extend, as a thin layer, similar finely striated masses, and these would have been on the top of the ciliary processes." He further states that these fibres may be traced to the anterior and posterior limb of the lens-capsule, and that amongst them are numerous cells. Kölliker confirms Lieberkühn's statements. There can be little doubt that the fibres of the zonula are of the nature of connective tissue: they are stated to be elastic. By Löwe they are believed to be developed out of the substance of the vitreous humour, but this does not appear to me to follow from the observations hitherto made. It seems quite possible that they arise from mesoblast cells which have grown into the cavity of the vitreous humour, solely in connection with their production.

The integral parts of the eye in front of the lens are the cornea, the aqueous humour, and the iris. The development of the latter has already been described, and there remain to be dealt with the cornea, and the cavity containing the aqueous humour.

The cornea. The cornea is formed by the coalescence of two structures, viz. the epithelium of the cornea and the cornea proper. The former is directly derived from the external epiblast, which covers the eye after the invagination of the lens. The latter is formed in a somewhat remarkable manner, first clearly made out by Kessler.

[FIG. 290. SECTION THROUGH THE EYE OF A FOWL ON THE EIGHTH DAY OF
DEVELOPMENT, TO SHEW THE IRIS AND CORNEA IN THE PROCESS OF
FORMATION. (After Kessler.)

_ep._ epiblastic epithelium of cornea; _cc._ corneal corpuscles
growing into the structureless matrix of the cornea; _dm._
Descemet's membrane; _ir._ iris; _cb._ mesoblast of the iris (this
reference letter points a little too high).

The space between the layers _dm._ and _ep._ is filled with the
structureless matrix of the cornea.]

When the lens is completely separated from the epidermis its outer wall is directly in contact with the external epiblast (future corneal epithelium). At its edge there is a small ring-shaped space bounded by the outer skin, the lens and the edge of the optic cup. In the chick, which we may take as typical, there appears at about the time when the cavity of the lens is completely obliterated a structureless layer external to the above ring-like space and immediately adjoining the inner face of the epiblast. This layer, which forms the commencement of the cornea proper, at first only forms a ring at the border of the lens, thickest at its outer edge, and gradually thinning off to nothing towards the centre. It soon however becomes broader, and finally forms a continuous stratum of considerable thickness, interposed between the external skin and the lens. As soon as this stratum has reached a certain thickness, a layer of flattened cells grows in along its inner side from the mesoblast surrounding the optic cup (fig. 290, _dm_). This layer is the epithelioid layer of the membrane of Descemet. After it[190] has become completely established, the mesoblast around the edge of the cornea becomes divided into two strata; an inner one (fig. 290, _cb_) destined to form the mesoblastic tissue of the iris already described, and an outer one (fig. 290, _cc_) adjoining the epidermis. The outer stratum gives rise to the corneal corpuscles, which are the only constituents of the cornea not yet developed. The corneal corpuscles make their way through the structureless corneal layer, and divide it into two strata, one adjoining the epiblast, and the other adjoining the inner epithelium. The two strata become gradually thinner as the corpuscles invade a larger and larger portion of their substance, and finally the outermost portion of them alone remains as the membrana elastica anterior and posterior (Descemet's membrane) of the cornea. The corneal corpuscles, which have grown in from the sides, thus form a layer which becomes continually thicker, and gives rise to the main substance of the cornea. Whether the increase in the thickness of the layer is due to the immigration of fresh corpuscles, or to the division of those already there, is not clear. After the cellular elements have made their way into the cornea, the latter becomes continuous at its edge with the mesoblast which forms the sclerotic.

[190] It appears to me possible that Lieberkühn may be right in
stating that the epithelium of Descemet's membrane grows in
between the lens and the epiblast before the formation of the
cornea proper, and that Kessler's account, given above, may on
this point require correction. From the structure of the eye in
the Ammocoete it seems probable that Descemet's membrane is
continuous with the choroid.

The derivation of the original structureless layer of the cornea is still uncertain. Kessler derives it from the epiblast, but it appears to me more probable that Kölliker is right in regarding it as derived from the mesoblast. The grounds for this view are, (1) the fact of its growth inwards from the border of the mesoblast round the edge of the eye, (2) the peculiar relations between it and the corneal corpuscles at a later period. This view would receive still further support if a layer of mesoblast between the lens and the epiblast were really present as believed by Lieberkühn. It must however be admitted that the objections to Kessler's view of its epiblastic nature are rather _a priori_ than founded on definite observation.

The observations of Kessler, which have been mainly followed in the above account, are strongly opposed by Lieberkühn (No. 374) and Arnold (No. 370), and are not entirely accepted by Kölliker. It is especially on the development of these parts in Mammalia (to be spoken of in the sequel) that the above authors found their objections. I have had through Kessler's kindness an opportunity of looking through some of his beautiful preparations, and have no hesitation in generally accepting his conclusions, though as mentioned above I cannot agree with all his interpretations.

The aqueous humour. The cavity for the aqueous humour has its origin in the ring-shaped space round the front of the lens, which, as already mentioned, is bounded by the external skin, the edge of the optic cup, and the lens. By the formation of the cornea this space is shut off from the external skin, and on the appearance of the epithelioid layer of Descemet's membrane a continuous cavity is developed between the cornea and the lens. This cavity enlarges and receives its final form on the full development of the iris.

_Comparative view of the development of the Vertebrate Eye._

The organ of vision, when not secondarily aborted, contains in all Vertebrata the essential parts above described. The most interesting cases of partial degeneration are those of Myxine and the Ammocoete. The development of such aborted eyes has as yet been studied only in the Ammocoete[191], in which it resembles in most important features that of other Vertebrata.

[191] The most detailed account is that of W. Müller (No. 377).

[FIG. 291. HORIZONTAL SECTION THROUGH THE HEAD OF A JUST HATCHED
LARVA OF PETROMYZON SHEWING THE DEVELOPMENT OF THE LENS OF THE EYE.

_th.c._ thalamencephalon; _op.v._ optic vesicle; _l._ lens of eye;
_h.c._ head cavity.]

Eye of Ammocoetes. The optic vesicle arises as an outgrowth of the fore-brain, but the secondary optic cup is remarkable in the young larva for its small size (fig. 291, _opv_). The thicker outer wall gives rise to the retina, and the thinner inner wall to the choroid pigment. The lens is formed as an invagination of the single-layered epidermis (fig. 291, _l_). As development proceeds the parts of the eye gradually enlarge, and the mesoblast around the hinder and dorsal part of the optic cup becomes pigmented. There is at first no cavity for the vitreous humour, but eventually the growth of the optic cup gives rise to a space, into which a cellular process of mesoblast grows at a slight notch in the ventral edge of the optic cup (W. Müller, No. 377). This notch is the only rudiment of the choroid fissure of other types. The mesoblastic process is probably the homologue of the processus falciformis and pecten, and appears to give rise to the vitreous humour; for a long time it retains its connection with the surrounding mesoblast. Its cells eventually disappear, and it never contains any vascular structures.

The lens for a long time remains as an oval vesicle with a central cavity. In a later stage, when the Ammocoete is fully developed, the secondary optic cup forms a deep pit (fig. 292, _r_); in the mouth of which is placed the lens (_l_). The two walls of the retina have now the normal vertebrate structure, though the pigment is as yet imperfectly present in the choroid layer. The lens has the embryonic forms of higher types (cf. fig. 289), consisting of an inner thicker segment, the true lens, and an outer layer forming the epithelium of the lens capsule. The edge of the optic cup, which forms the rudiment of the epiblast of the iris, is imperfectly separated from the remainder of the optic cup; and a mesoblastic element of the iris, distinct from Descemet's membrane (_dm_), can hardly be spoken of.

There is no cavity for the aqueous humour in front of the lens; and there is no cornea as distinct from the epidermis and subepidermic tissues. The elements in front of the lens are (1) the epidermis (_ep_); (2) the dermis (_dc_); (3) the subdermal connective tissue (_sdc_) which passes without any sharp line of demarcation into the dermis; (4) a thick membrane, continuous with the mesoblastic part of the choroid, which appears to represent Descemet's membrane. The subdermal connective tissue is continued as an investment round the whole eye; and there is no differentiated sclerotic and only an imperfect choroid.

In a still later stage a distinct mesoblastic element for the iris is formed. When the Ammocoete is becoming a Lamprey, the eye approaches the surface; an anterior chamber is established; and the eye differs from that of the higher types mainly in the fact that the cornea is hardly distinguished from the remainder of the skin, and that a sclerotic is very imperfectly represented.

Optic vesicles. The development of the primitive optic vesicles, so far as is known, is very constant throughout the Vertebrata. In Teleostei and Lepidosteus alone is there an important deviation from the ordinary type, dependent however upon the mode of formation of the medullary keel, the optic vesicles arising while the medullary keel is still solid, and being at first also solid. They subsequently acquire a lumen and undergo the ordinary changes.

[FIG. 292. EYE OF AN AMMOCOETES LYING BENEATH THE SKIN.

_ep._ epidermis; _d.c._ dermal connective tissue continuous with the
subdermal connective tissue (_s.d.c_), which is also shaded. There
is no definite boundary to this tissue where it surrounds the eye.

_m._ muscles; _dm._ membrane of Descemet; _l._ lens; _v.h._ vitreous
humour; _r._ retina; _rp._ retinal pigment.]

The lens. In the majority of groups, viz. Elasmobranchii, Reptilia, Aves, and Mammalia, the lens is formed by an open invagination of the epiblast, but in Amphibia, Teleostei and Lepidosteus, where the nervous layer of the skin is early established, this layer alone takes part in the formation of the lens (fig. 293, _l_). The lens is however formed even in these types as a hollow body by an invagination; but its opening remains permanently shut off from communication with the exterior by the epidermic layer of the epiblast. Götte describes the lens as formed by a solid thickening of the nervous layer in Bombinator. This is probably a mistake.

The cornea. The mode of formation of the cornea already described appears to be characteristic of most Vertebrata except the Ammocoete. It has been found by Kessler in Aves, Reptilia and Amphibia, and probably also occurs in Pisces. In Mammals it is not however so easy to establish. There are at first no mesoblast cells between the lens and the epiblast (fig. 295) but in many Mammals (_vide_ Kessler, No. 372, pp. 91-94) a layer of rounded mesoblast cells, which forms Descemet's membrane, grows in between the two, at a time when it is not easy to recognise a corneal lamina, as distinct from a simple coagulum.

After the formation of this layer the mesoblast cells grow into the corneal lamina from the sides, and becoming flattened arrange themselves in rows between the laminæ of the cornea. The cornea continues to increase in thickness by the addition of laminæ on the side adjoining the epiblast.

We have already seen that in the Lamprey the cornea is nothing else but the slightly modified and more transparent epidermis and dermis.

The optic nerve and the choroid fissure. It will be convenient to consider together the above structures, and with them the vascular and other processes which pass into the cavity of the optic cup through the choroid fissure. These parts present on the whole a greater amount of variation than any other parts of the eye.

I commence with the Fowl which is both a very convenient general type for comparison, and also that in which these structures have been most fully worked out.

During the third day of incubation there passes in through the choroid slit a vascular loop, which no doubt supplies the transuded material for the growth of the vitreous humour. Up to the fifth day this vascular loop is the only structure passing through the choroid slit. On this day however a new structure appears, which remains permanently through life, and is known as the pecten. It consists of a lamellar process of the mesoblast cells round the eye, passing through the choroid slit near the optic nerve, and enveloping part of the afferent branch of the vascular loop above mentioned. The proximal part of the free edge of the pecten is somewhat swollen, and sections through this part have a club-shaped form. On the sixth day the choroid slit becomes rapidly closed, so that at the end of the sixth day it is reduced to a mere seam. There are however two parts of this seam where the edges of the optic cup have not coalesced. The proximal of these adjoins the optic nerve, and permits the passage of the pecten and at a later period of the optic nerve; and the second or distal one is placed near the ciliary edge of the slit, and is traversed by the efferent branch of the above-mentioned vascular loop. This vessel soon atrophies, and with it the distal opening in the choroid slit completely vanishes. In some varieties of domestic Fowl (Lieberkühn) the opening however persists. The seam which marks the original site of the choroid slit is at first conspicuous by the absence of pigment, and at a later period by the deep colour of its pigment. Finally, a little after the ninth day, no trace of it is to be seen.

[FIG. 293. SECTION THROUGH THE FRONT PART OF THE HEAD OF A
LEPIDOSTEUS EMBRYO ON THE SEVENTH DAY AFTER IMPREGNATION.

_al._ alimentary tract; _fb._ thalamencephalon; _l._ lens of eye;
_op.v._ optic vesicle. The mesoblast is not represented.]

Up to the eighth day the pecten remains as a simple lamina; by the tenth or twelfth day it begins to be folded or rather puckered, and by the seventeenth or eighteenth day it is richly pigmented and the puckerings have become nearly as numerous as in the adult, there being in all seventeen or eighteen. The pecten is almost entirely composed of vascular coils, which are supported by a sparse pigmented connective tissue; and in the adult the pecten is still extremely vascular. The original artery which became enveloped at the formation of the pecten continues, when the latter becomes vascular, to supply it with blood. The vein is practically a fresh development after the atrophy of the distal portion of the primitive vascular loop of the vitreous humour.

There are no true retinal blood-vessels.

In the formation of the optic cup the extreme peripheral part of the optic nerve, which is in immediate proximity with the artery of the pecten, becomes folded. The permanent opening in the choroid fissure for the pecten is intimately related to the entrance of the optic nerve into the eyeball; the fibres of the optic nerve passing in at the inner border of the pecten, coursing along its sides to its outer border, and radiating from it as from a centre to all parts of the retina.

In the Lizard the choroid slit closes considerably earlier than in the Fowl. The vascular loop in the vitreous humour is however more developed. The pecten long remains without vessels, and does not in fact become at all vascular till after the very late disappearance of the distal part of the vascular loop of the vitreous humour.

The arrangement of the ingrowth through the choroid slit in Elasmobranchii (Scyllium) has been partially worked out, and so far as is at present known the agreement between the Avian and Elasmobranch type is fairly close.

At the time when the cavity between the lens and the secondary optic cup is just commencing to be formed, a process of mesoblast accompanied by a vascular loop passes into the vitreous humour, through the choroid slit, close to the optic nerve. The vessel in this process is no doubt equivalent to the vascular loop in the Avian eye, but I have not made out that it projects beyond the mesoblastic process accompanying it. As the cavity of the vitreous humour enlarges and the choroid slit elongates, the process through it takes the form of a lamina with a somewhat swollen border, and projects for some distance into the cavity of the vitreous humour.

At a later stage, after the outer layer of the optic cup has become pigmented, the distal part of the choroid slit adjoining the border of the lens closes up; but along the line where it was present the walls of the optic cup remain very thin and are thrown into three folds, two lateral and one median, projecting into the cavity of the vitreous humour. The median fold is in contact with the lens, and the vascular mesoblast surrounding the eye projects into the space between the two laminæ of which it is formed. In passing from the region of the lens to that of the optic nerve the lateral folds of the optic cup disappear, and the median fold forms a considerable projection into the cavity of the vitreous humour. It consists of a core of mesoblast covered by a delicate layer derived from both strata of the optic cup. Still nearer the optic nerve the choroid slit is no longer closed, and the mesoblast, which in the neighbourhood of the lens only extended into the folds of the wall of the optic cup, now projects freely into the cavity of the vitreous humour, and forms the lamina already described. It is not very vascular, but close to the optic nerve there passes into it a considerable artery.

In the young animal the choroid slit is no longer perforated by a mesoblastic lamina. At its inner end it remains open to allow of the passage of the optic nerve. The line of the slit can easily be traced along the lower side of the retina; and close to the lens the retinal wall continues, as in the embryo, to be raised into a projecting fold. Traces of these structures are visible even in the fully grown examples of Scyllium.

As has been pointed out by Bergmeister the mesoblastic lamina projecting into the vitreous humour resembles the pecten at an early stage of development, and is without doubt homologous with it. The artery which supplies it is certainly equivalent to the artery of the pecten.

There can be no doubt that the mesoblastic lamina projecting into the vitreous humour is equivalent to the processus falciformis of Teleostei, and it seems probable that the whole of it, including the free part as well as that covered by epiblast, ought to be spoken of under this title. The optic nerve in Elasmobranchii is not included in the folding to which the secondary optic vesicle owes its origin, and would seem to perforate the walls of the optic cup only at the distal end of the processus falciformis.

[FIG. 294. HORIZONTAL SECTION THROUGH THE EYE OF A TELEOSTEAN
EMBRYO. (From Gegenbaur; after Schenk.)

_s._ choroid fissure, with two folds forming part of the processus
falciformis; _a._ choroid layer of optic cup; _b._ retinal layer of
optic cup; _c._ cavity of vitreous humour; _d._ lens.]

In Teleostei there is at first a vascular loop like that in Birds, passing through the choroid fissure. This has been noticed by Kessler in the Pike, and by Schenk in the Trout. At a later period a mesoblastic ingrowth with a blood-vessel makes its way in many forms into the cavity of the vitreous humour, accompanied by two folds in the walls of the free edges of the choroid fissure (fig. 294). These structures, which constitute the processus falciformis, clearly resemble very closely the mesoblastic process and folds of the optic cup in Elasmobranchii. The processus falciformis comes in contact with, and perhaps becomes attached to the wall of the lens; and persists through life.

In Triton there is no vascular ingrowth through the choroid fissure, but a few mesoblastic cells pass in which represent the vascular ingrowth of other types. The optic nerve perforates the proximal extremity of the original choroid slit.

The absence of an embryonic blood-vessel does not however hold good for all Amphibia, as there is present in the embryo Alytes (Lieberkühn) an artery, which breaks up into a capillary system on the retinal border of the vitreous humour.

In the Ammocoete the choroid slit is merely represented by a slight notch on the ventral edge of the optic cup, and the mesoblastic process which passes through the choroid slit in most types is represented by a large cellular process, from which the vitreous humour would appear to be derived.

Mammalia differ from all the types already described in the immense foetal development of the blood-vessels of the vitreous humour. There are however some points in connection with the development of these vessels which are still uncertain. The most important of these points concerns the presence of a prolongation of the mesoblast around the eye into the cavity of the vitreous humour. It is maintained by Lieberkühn, Arnold, Kölliker, etc., that in the invagination of the lens a thin layer of mesoblast is carried before it; and is thus transported into the cavity of the vitreous humour. This is denied by Kessler, but the layer is so clearly figured by the above embryologists, that the existence of it in some Mammalia (the Rabbit, etc.) must I think be accepted.

In the folding in of the optic vesicle, which accompanies the formation of the lens, the optic nerve becomes included, and on the development of the cavity of the vitreous humour an artery, running in the fold of the optic nerve, passes through the choroid slit into the cavity of the vitreous humour (fig. 295, _acr_). The sides of the optic nerve subsequently bend over, and completely envelope this artery, which at a later period gives off branches to the retina, and becomes known as the arteria centralis retinæ. It is homologous with the arterial limb of the vascular loop projecting into the vitreous humour in Birds, Lizards, Teleostei, etc.

[FIG. 295. SECTION THROUGH THE EYE OF A RABBIT EMBRYO OF ABOUT
TWELVE DAYS..

_c._ epithelium of cornea; _l._ lens; _mec._ mesoblast growing in
from the side to form the cornea; _rt._ retina; _a.c.r._ arteria
centralis retinæ; _of.n._ optic nerve.

The figure shews (1) the absence at this stage of mesoblast between the lens and the epiblast: the interval between the two has however been made too great; (2) the arteria centralis retinæ forming the vascular capsule of the lens and continuous with vascular structures round the edges of the optic cup.]

Before becoming enveloped in the optic nerve this artery is continued through the vitreous humour (fig. 295), and when it comes in close proximity to the lens it divides into a number of radiating branches, which pass round the edge of the lens, and form a vascular sheath which is prolonged so as to cover the anterior wall of the lens. In front of the lens they anastomose with vessels, coming from the iris, many of which are venous (fig. 295)--and the whole of the blood from the arteria centralis is carried away by these veins. The vascular sheath surrounding the lens receives the name of the membrana capsulo-pupillaris. The posterior part of it appears (Kessler, No. 372) to be formed of vessels without the addition of any other structures and is either formed simply by branches of the arteria centralis, or out of the mesoblast cells involuted with the lens. The anterior part of the vascular sheath is however inclosed in a very delicate membrane, the membrana pupillaris, continuous at the sides with the epithelium of Descemet's membrane. On the formation of the iris this membrane lies superficially to it, and forms a kind of continuation of the mesoblast of the iris over the front of the lens.

The origin of this membrane is much disputed. By Kessler, whose statements have been in the main followed, it is believed to appear comparatively late as an ingrowth of the stroma of the iris; while Kölliker believes it to be derived from a mesoblastic ingrowth between the front wall of the lens and the epiblast. According to Kölliker this ingrowth subsequently becomes split into two laminæ, one of which forms the cornea, and the other the anterior part of the vascular sheath of the lens with its membrana pupillaris. Between the two appears the aqueous humour.

The membrana capsulo-pupillaris is simply a provisional embryonic structure, subserving the nutrition of the lens. The time of its disappearance varies somewhat for the different Mammalia in which this point has been investigated. In the human embryo it lasts from the second to the seventh month and sometimes longer. As a rule it is completely absorbed at the time of birth. The absorption of the anterior part commences in the centre and proceeds outwards.

In addition to the vessels of the vascular capsule round the lens, there arise from the arteria centralis retinæ, just after its exit from the optic nerve, in many forms (Dog, Cat, Calf, Sheep, Rabbit, Man) provisional vascular branches which extend themselves in the posterior part of the vitreous humour. Near the ciliary end of the vitreous humour they anastomose with the vessels of the membrana capsulo-pupillaris.

In Mammals the choroid slit closes very early, and is not perforated by any structure homologous with the pecten. The only part of the slit which remains open is that perforated by the optic nerve; and in the centre of the latter is situated the arteria centralis retinæ as explained above. From this artery there grow out the vessels to supply the retina, which have however nothing to do with the provisional vessels of the vitreous humour just described (Kessler). On the atrophy of the provisional vessels the whole of the blood of the arteria centralis passes into the retina.

It is interesting to notice (Kessler, No. 372, p. 78) that there seems to be a blood-vessel supplying the vitreous humour in the embryos of nearly all vertebrate types, which is homologous throughout the Vertebrata. This vessel often exhibits a persisting and a provisional part. The latter in Mammalia is the membrana capsulo-pupillaris and other vessels of the vitreous humour; in Birds and Lizards it is the part of the original vascular loop, not included in the pecten, and in Osseous Fishes that part (?) not involved in the processus falciformis. The permanent part is formed by the retinal vessels of Mammalia, by the vessels of the pecten in Birds and Lizards, and by those of the processus falciformis in Fishes.

The Iris and Ciliary processes. The walls of the edge of the optic cup become very much thinner than those of the true retinal part. In many Vertebrates (Mammalia, Aves, Reptilia, Elasmobranchii, etc.) the thinner part, together with the mesoblast covering it, becomes divided into two regions, viz. that of the iris, and that of the ciliary processes. In the Newt and Lamprey this differentiation does not take place, but the part in question simply becomes the iris.

_Accessory Organs connected with the Eye._

Eyelids. The most important accessory structures connected with the eye are the eyelids. They are developed as simple folds of the integument with a mesoblastic prolongation between their two laminæ. They may be three in number, viz. an upper and lower, and a lateral one--the nictitating membrane--springing from the inner or anterior border of the eye. Their inner face is lined by a prolongation of conjunctiva, which is the modified epiblast covering the cornea and part of the sclerotic.

In Teleostei and Ganoidei eyelids are either not present or at most very rudimentary. In Elasmobranchii they are better developed, and the nictitating membrane is frequently present. The latter is also usually found in Amphibia. In the Sauropsida all three eyelids are usually present, but in Mammalia the nictitating membrane is rudimentary.

In many Mammalia the two eyelids meet together during a period of embryonic life, and unite in front of the eye. A similar arrangement is permanent through life in Ophidia and some Lacertilia; and there is a chamber formed between the coalesced eyelids and the surface of the cornea, into which the lacrymal ducts open.

Lacrymal glands. Lacrymal glands are found in the Sauropsida and Mammalia. They arise (Remak, Kölliker) as solid ingrowths of the conjunctival epithelium. They appear in the chick on the eighth day.

Lacrymal duct. The lacrymal duct first appears in Amphibia, and is present in all the higher Vertebrates. Its mode of development in the Amphibia, Lacertilia and Aves has recently been very thoroughly worked out by Born (Nos. 380 and 381).

In Amphibia he finds that the lacrymal duct arises as a solid ridge of the mucous layer of the epidermis, continued from the external opening of the nasal cavity backwards towards the eye. It usually appears at about the time when the nasal capsule is beginning to be chondrified. As this ridge is gradually prolonged backwards towards the eye its anterior end becomes separated from the epidermis, and grows inwards in the mesoblast to become continuous with the posterior part of the nasal sack. The posterior end which joins the eye becomes divided into the two collecting branches of the adult. Finally the whole structure becomes separated from the skin except at the external opening, and develops a lumen.

In Lacertilia the lacrymal duct arises very much in the same manner as in Amphibia, though its subsequent growth is somewhat different. It appears as an internal ridge of the epithelium, at the junction of the superior maxillary process and the fold which gives rise to the lower eyelid. A solid process of this ridge makes its way through the mesoblast on the upper border of the maxillary process till it meets the wall of the nasal cavity, with the epithelium of which it becomes continuous. At a subsequent stage a second solid growth from the upper part of the epithelial ridge makes its way through the lower eyelid, and unites with the inner epithelium of the eyelid; and at a still later date a third growth from the lower part of the structure forms a second junction with the epithelium of the eyelid. The two latter outgrowths form the two upper branches of the duct. The ridge now loses its connection with the external skin, and, becoming hollow, forms the lacrymal duct. It opens at two points on the inner surface of the eyelid, and terminates at its opposite extremity by opening into the nasal cavity. It is remarkable, as pointed out by Born, that the original epithelial ridge gives rise directly to a comparatively small part of the whole duct.

In the Fowl the lacrymal duct is formed as a solid ridge of the epidermis, extending along the line of the so-called lacrymal groove from the eye to the nasal pit (fig. 120). At the end of the sixth day it begins to be separated from the epidermis, remaining however united with it on the inner side of the lower eyelid. After its separation from the epidermis it forms a solid cord, the lower end of which unites with the wall of the nasal cavity. The cord so formed gives rise to the whole of the duct proper and to the lower branch of the collecting tube. The upper branch of the collecting tube is formed as an outgrowth from this cord. A lumen begins to be formed on the twelfth day of incubation, and first appears at the nasal end. It arises by the formation of a space between the cells of the cord, and not by an absorption of the central cells.

In Mammalia Kölliker states that he has been unable to observe anything similar to that described by Born in the Sauropsida and Amphibia, and holds to the old view, originally put forward by Coste, that the duct is formed by the closure of a groove leading from the eye to the nose between the outer nasal process and the superior maxillary process. The upper extremity of the duct dilates to form a sack, from which two branches pass off to open on the lacrymal papillæ. In view of Born's discoveries Kölliker's statements must be received with some caution.

_The Eye of the Tunicata._

The unpaired eye of the larva of simple Ascidians is situated somewhat to the right side of the posterior part of the dorsal wall of the anterior cephalic vesicle (fig. 296, _O_). It consists of a refractive portion, turned towards the cavity of the vesicle of the brain, and a retinal portion forming part of the wall of the brain. The refractive parts consist of a convex-concave meniscus in front, and a spherical lens behind, adjoining the concave side of the meniscus. The posterior part of this lens is imbedded in a layer of pigment. The retina is formed of columnar cells, with their inner ends imbedded in the pigment which encloses the posterior part of the lens. The retinal part of the eye arises in the first instance as a prominence of the wall of the cerebral vesicle: its cells become very columnar and pigmented at their inner extremities (fig. 8, _V_, _a_). The lens is developed at a later period, after the larva has become hatched, but the mode of its formation has not been made out.

[FIG. 296. LARVA OF ASCIDIA MENTULA. (From Gegenbaur; after
Kupffer.) Only the anterior part of the tail is represented.

_N´._ anterior swelling of neural tube; _N._ anterior swelling of
spinal portion of neural tube; _n._ hinder part of neural tube;
_ch._ notochord; _K._ branchial region of alimentary tract; _d._
oesophageal and gastric region of alimentary tract; _O._ eye; _a._
otolith; _o._ mouth; _s._ papilla for attachment.]

_General considerations on the Eye of the Chordata._

There can be but little doubt that the eye of the Tunicata belongs to the same phylum as that of the true Vertebrata, different as the two eyes are. The same may also be said with reference to the degenerate and very rudimentary eye of Amphioxus.

The peculiarity of the eye of all the Chordata consists in the retina being developed from part of the wall of the brain. How is this remarkable feature of the eye of the Chordata to be explained?

Lankester, interpreting the eye in the light of the Tunicata, has made the interesting suggestion[192] "that the original Vertebrate must have been a transparent animal, and had an eye or pair of eyes inside the brain, like that of the Ascidian Tadpole."

[192] _Degeneration_, London, 1880, p. 49.

This explanation may possibly be correct, but another explanation appears to me possible, and I am inclined to think that the vertebrate eyes have not been derived from eyes like those of Ascidians, but that the latter is a degenerate form of vertebrate eye.

The fact of the retina being derived from the fore-brain may perhaps be explained in the same way as has already been attempted in the case of the retina of the Crustacea; _i.e._ by supposing that the eye was evolved simultaneously with the fore part of the brain.

The peculiar processes which occur in the formation of the optic vesicle are more difficult to elucidate; and I can only suggest that the development of a primary optic vesicle, and its conversion into an optic cup, is due to the retinal part of the eye having been involved in the infolding which gave rise to the canal of the central nervous system. The position of the rods and cones on the posterior side of the retina is satisfactorily explained by this hypothesis, because, as may be easily seen from figure 285, the posterior face of the retina is the original external surface of the epidermis, which is infolded in the formation of the brain; so that the rods and cones are, as might be anticipated, situated on what is morphologically the external surface of the epiblast of the retina.

The difficulty of this view arises in attempting to make out how the eye can have continued to be employed during the gradual change of position which the retina must have undergone in being infolded with the brain in the manner suggested. If however the successive steps in this infolding were sufficiently small, it seems to me not impossible that the eye might have continued to be used throughout the whole period of change, and a transparency of the tissues, such as Lankester suggests, may have assisted in rendering this possible.

The difficulty of the eye continuing to be in use when undergoing striking changes in form is also involved in Lankester's view, in that if, as I suppose, he starts from the eye of the Ascidian Tadpole with its lenses turned _towards_ the cavity of the brain; it is necessary for him to admit that a fresh lens and other optical parts of the eye became developed on the _opposite side_ of the eye to the original lens; and it is difficult to understand such a change, unless we can believe that the refractive media on the two sides were in operation simultaneously. It may be noted that the same difficulty is involved in supposing, as I have done, that the eye of the Ascidian Tadpole was developed from that of a Vertebrate. I should however be inclined to suggest that the eye had in this case ceased for a period to be employed; and that it has been re-developed again in some of the larval forms. Its characters in the Tunicata are by no means constant.

_Accessory eyes in the Vertebrata._

In addition to the paired eyes of the Vertebrata certain organs are found in the skin of a few Teleostei living in very deep water, which, though clearly not organs of true vision, yet present characters which indicate that they may be used in the perception of light. The most important of such organs are those found in Chauliodus, Stomias, etc., the significance of which was first pointed out by Leuckart, while the details of their structure have been recently worked out by Leydig[193] and Ussow. They are distributed not only in the skin, but are also present in the mouth and respiratory cavity, a fact which appears to indicate that their main function must be something else than the perception of light. It has been suggested that they have the function of producing phosphorescence.

Another organ, probably of the same nature, is found on the head of Scopelus.

The organs in Chauliodus are spherical or nearly spherical bodies invested in a special tunic. The larger of them, which alone can have any relation to vision, are covered with pigment except on their outer surface. The interior is filled with two masses, named by Leuckart the lens and vitreous humour. According to Leydig each of them is cellular and receives a nerve, the ultimate destination of which has not however been made out. According to Ussow the anterior mass is structureless, but serves to support a lens, placed in the centre of the eye, and formed of a series of crystalline cones prolonged into fibres, which in the posterior part of the eye diverge and terminate by uniting with the processes of multipolar cells, placed near the pigmented sheath. These cells, together with the fibres of the crystalline cones which pass to them, are held by Ussow to constitute a retina.

_Eye of the Mollusca._

(362) N. Bobretzky. "Observations on the development of the Cephalopoda" (Russian). _Nachrichten d. kaiserlichen Gesell. d. Freunde der Naturwiss. Anthropolog. Ethnogr. bei d. Universität Moskau._

(363) H. Grenacher. "Zur Entwicklungsgeschichte d. Cephalopoden." _Zeit. f. wiss. Zool._, Bd. XXIV. 1874.

(364) V. Hensen. "Ueber d. Auge einiger Cephalopoden." _Zeit. f. wiss. Zool._, Vol. XV. 1865.

(365) E. R. Lankester. "Observations on the development of the Cephalopoda." _Quart. J. of Micr. Science_, Vol. XV. 1875.

(366) C. Semper. _Ueber Sehorgane von Typus d. Wirbelthieraugen._ Wiesbaden, 1877.

_Eye of the Arthropoda._

(367) N. Bobretzky. _Development of Astacus and Palaemon._ Kiew, 1873.

(368) A. Dohrn. "Untersuchungen üb. Bau u. Entwicklung d. Arthropoden. Palinurus und Scyllarus." _Zeit. f. wiss. Zool._, Bd. XX. 1870, p. 264 et seq.

(369) E. Claparède. "Morphologie d. zusammengesetzten Auges bei den Arthropoden." _Zeit. f. wiss. Zool._, Bd. X. 1860.

(370) H. Grenacher. _Untersuchungen üb. d. Sehorgane d. Arthropoden._ Göttingen, 1879.

_Vertebrate Eye._

(371) J. Arnold. _Beiträge zur Entwicklungsgeschichte des Auges._ Heidelberg, 1874.

(372) Babuchin. "Beiträge zur Entwicklungsgeschichte des Auges." _Würzburger naturwissenschaftliche Zeitschrift_, Bd. 8.

(373) L. Kessler. _Zur Entwicklung d. Auges d. Wirbelthiere._ Leipzig, 1877.

(374) N. Lieberkühn. _Ueber das Auge des Wirbelthierembryo._ Cassel, 1872.

(375) N. Lieberkühn. "Beiträge z. Anat. d. embryonalen Auges." _Archiv f. Anat. und Phys._, 1879.

(376) L. Löwe. "Beiträge zur Anatomie des Auges" and "Die Histogenese der Retina." _Archiv f. mikr. Anat._, Vol. XV. 1878.

(377) V. Mihalkowics. "Untersuchungen über den Kamm des Vogelauges." _Archiv f. mikr. Anat._, Vol. IX. 1873.

(378) W. Müller. "Ueber die Stammesentwickelung des Schorgans der Wirbelthiere." _Festgabe Carl Ludwig._ Leipzig, 1874.

(379) S. L. Schenk. "Zur Entwickelungsgeschichte des Auges der Fische." _Wiener Sitzungsberichte_, Bd. LV. 1867.

_Accessory organs of the Vertebrate Eye._

(380) G. Born. "Die Nasenhöhlen u. d. Thränennasengang d. Amphibien." _Morphologisches Jahrbuch_, Bd. II. 1876.

(381) G. Born. "Die Nasenhöhlen u. d. Thränennasengang d. amnioten Wirbelthiere. I. Lacertilia. II. Aves." _Morphologisches Jahrbuch_, Bd. V. 1879.

_Eye of the Tunicata._

(382) A. Kowalevsky. "Weitere Studien üb. d. Entwicklung d. einfachen Ascidien." _Archiv f. mikr. Anat._, Vol. VII. 1871.

(383) C. Kupffer. "Zur Entwicklung d. einfachen Ascidien." _Archiv f. mikr. Anat._, Vol. VII. 1872.

[193] F. Leydig. "Ueber Nebenaugen d. Chauliodus Sloani."
_Archiv f. Anat. und Phys._, 1879. M. Ussow. "Ueb. d. Bau d.
augenähnlichen Flicken einiger Knochenfische." _Bul. d. la Soc.
d. Naturalistes de Moscou_, Vol. LIV. 1879.
_Vide_ for general description and further literature,
Günther, _The Study of Fishes_, Edinburgh, 1880.