On the Structure and Physiology of the Retina

In his general account of the structure of the retina we do not perceive that Professor Schultze differs very materially from most later writers on the subject. What he says respecting it may, however, be very briefly stated as follows :

The retina in man is composed of a fibrous or trabecular framework, composed of connective tissue, and which serves as a support to the nervous or sentient elements. The fibrous framework consists of an outer and an inner membrana limitons, connected together by a network of fibres, the principal of which, passing from one limiting membrane to the other, constitute the “radial fibres of Millier.” These are connected by irregular lateral fibres, so that the whole constitutes, speaking generally, a sort of wide trabecular network; but at two special levels in the retina the fibrous tissue forms a very close, almost membraniform plexus, the outer and thinner of which corresponds with the so-termed “intergranular layer,” and the inner in the same manner corresponds with the “molecular layer “or outer part of the “layer of grey substance.” The membrana limitans externa in the fully developed organ does not constitute a continuous expansion, but is perforated with numerous closely placed openings, like the shelf of a bottle-rack. The membrana limitans interna, properly speaking, is also not a continuous membrane, but a reticulated tissue composed of the expanded ends of the radial trabeculæ or “fibres of Müller.” This fibrous framework supports the nervous part of the retina, which may be subdivided into six, or more properly, perhaps, seven distinct layers. These layers, proceeding from without inwards, are—1. The bacillary layer, composed of “rods “and “cones,” placed vertically on the periphery, and each lodged by its inner extremity in one of the openings in the outer limitary membrane. 2. The “outer granule-layer,” composed for the most part of granular nucleated cells, connected with either the “rods “or “cones,” and traversed by the filaments proceeding from those bodies. 3. The “intergranular layer,” which is constituted, as before remarked, in part of a fine, fibrous, trabecular network, intermixed with which is a still finer plexus of very delicate nerve-fibres, for the most part, as it would seem, continuous with the terminal fibrillæ of the cone-filaments, and perhaps also in part with the terminations of the rod-filaments, although this has not been as yet clearly made out. 4. The inner granule-layer, containing for the most part bipolar ganglion-cells and abundance of fine nerve-filaments. 5. The “molecular layer,” which is of considerable thickness, and, like the “intergranular layer,” apparently composed of an intricate interlacement of very delicate nerve-filaments and the fine trabecular network before mentioned. 6. The “ganglionic layer,” constituted chiefly of large multipolar nerve-cells, each of which on its inner aspect appears to be connected with a fibrilla of the optic nerve, and on its outer to give off several processes which break up into the delicate fibrils contained in the molecular layer. 7. The layer of “optic nerve-fibres,” which in most animals appear to have no sheath, but to represent axial filaments.

The author’s researches have been directed more especially to the distinction between the “rods “and “cones.” But his attention has been turned, not so much to their morphological characters, with respect to which little now remains to be said, as to their relations to the other retinal elements, so that he might be able, if possible, to obtain some insight into their physiological differences. That such differences must exist cannot be doubted by any one who regards the unequal distribution of the two elements in different parts of the human retina, and remembers that in the most sensitive part of it, as is well known, “cones “only exist, whilst in every other part the “rods “far exceed the “cones “in number. But these conditions have hitherto remained unexplained, as has also the remarkable fact that in the retina of many animals the “rods “alone are found, and in others only “cones.” In the prosecution of his object, therefore, M. Schultze has found it necessary to examine, not only the human retina in its various regions, and particularly in the macula lútea and fovea centralis, but also to investigate all the varieties of structure exhibited in other animals. And in order to leave no means untried for arriving at a satisfactory elucidation of the subject, he has further closely studied the development of the retina, and particularly that of the bacillary layer.

The first section of the paper is devoted to the consideration mainly of the bacillary layer in the human subject, whose general structure is described much in the usual terms. The observations were made upon the recent human retina prepared with dilute osmic acid, and the beautiful illustrative figures are stated to have been taken from nature. They are excellently done, and doubtless accurately represent the structure as thus prepared. Retinas hardened by immersion in solutions of osmic acid containing per cent, are readily split up by means of needles into their laminæ parallel with the radial fibres ; and these products of natural fissure are clearly, the author thinks, preferable to thin sections. The principal points to which we shall refer, contained in this section, are :—(1) The fine longitudinal striation observable in the “cones “and “cone-filaments.” (2) That the space between the “cone-filaments,” as they cross the outer granule-layer, is entirely occupied by small, closely crowded cells, all of which are connected by finer or coarser filaments with the “rods.” These cells may be regarded, with H. Müller, as bipolar ganglion-cells. (3) The distinctive characteristics of “cone-filaments,” which are much thicker than those of the “rods,” are then detailed, and the differences between them and the fibrous radial trabeculæ pointed out.

The relations of the “rods “and “cones,” and the disposition of their filaments in the neighbour hood of and in the macula lútea, are next described, and particular pains are taken to render the structure of the retina in the macula and fovea centralis clear and intelligible, and, as it appears to us, with complete success.

The structure of the retina in mammals and other vertebrates is then compared with that of the human eye regarded as a typical form.

Apes, as is well known, possess a macula lútea, and in other respects their retina seems to agree very closely with that of man, even in the comparatively great thickness of the “cone-fibres.”

Among the other mammalia a very remarkable and, as it would seem, hitherto unnoticed diversity, with respect to the distribution of “rods “and “cones,” exists. Whilst most of our larger domestic animals, especially the sheep, ox, pig, horse, and dog, present an arrangement of those elements resembling that which is observed in the human subject and in apes, except, of course, in the absence of the macula lútea, the cones, according to the author’s observations, are entirely wanting in bats, the hedgehog, mole, mouse, and guinea-pig. A sort of intermediate condition is met with in the cat, rabbit, and rat, in which animals are found either very slender true “cones,” as in the cat, or merely indications of them, as in the rabbit. But in any case the “rods “preponderate so much that the “cones “among them may readily be over-looked. According to Ritter, the “cones “are also wanting in Balænfl mysticetus.

In the rat the “rods “are the longest and slenderest yet met with by the author. In the other vertebrate classes the proportion of “rods “and “cones “to each other approaches nearest to that observed in the mammalian retina in the osseous fishes. In the ray and shark “rods “only exist. In Petromyzon elements of one kind only occur in the bacillary layer ; but whether these be “cones “or ^rf rods “is undetermined, nor is it determined whether, as supposed by some, both elements may not really be present. The osseous fishes affotd excellent materials for the study of the “cone ‘‘-fibres ; which at one time M. Schultze regarded as belonging to the connective-tissue framework of the retina, and to represent in the outer granule-layer the “radial fibres of Muller “in the other layers of the retina ; but of their nervous nature, as of the corresponding fibres in the human retina, he is now thoroughly convinced.

The structure of the retina in birds, reptiles, and amphibia, differs in a very peculiar manner from that of mammals and fish. In the bird’s retina the proportion of “cones “to “rods “is in the reverse proportion to that in the mammalia. In other words,theretina of the bird, as regards the distribution of “rods “and “cones,” approaches that which is observed in the human macula lútea, inasmuch as the “cones “preponderate greatly over the “rods.” The same disposition is found in the retina of reptiles. In the turtle the arrangement is precisely the same as in birds, whilst in the lizards the “rods “are wholly wanting, as they would appear to be also in snakes. An exception, however, to this rule, as regards birds, is afforded in the owl, in several species of which (S. aluco, noctua, nnàflammea) the preponderance in number would seem to be in favour of the “rods and from this circumstance, as well as owing to the enormous length of the “.rods” in proportion to the “cones,” the mosaic aspect of the outer surface of the retina in these birds bears a striking resemblance to that of the bat. And owing to the same condition also, the owl’s retina is almost everywhere destitute of the colours so characteristic of the membrane in other birds. And another remarkable circumstance with respect to the retina in owls is the total absence in it of red pigment-globules ; and even the few yellow cones become paler and paler towards the ora serrata, until at length they are entirely colourless. These facts would seem to point out that, as the retina of nocturnal mammalia is distinguished by the total absence of “cones,” so in the case of the owl the comparative paucity of the same elements, together with the pale colour of the few pigment-globules, may also be connected with its nocturnal habits and avoidance of light. It would, therefore, M. Schultze remarks, be very interesting to examine the retina of other nocturnal birds, as of the Caprimulgidæ, &c.

Another and most characteristic peculiarity of the retina of birds, some reptiles and amphibia, but more especially of the first, is the presence in most of the “cones” of a spherical globule of red or yellow colour, but chiefly yellow, and which is situated at the junction of the inner and outer segments, that is to say, at the internal end of the latter, whose whole diameter is occupied by it, and consequently all the light reaching the outer segment of the cone must pass through this coloured medium. The author’s observations would seem to show that the yellow colour predominates in the more sensitive parts of the retina. At least, this presumption arises from the circumstance that in such birds as the pigeon, crow, and hawk (although swift-flying birds), which present a fovea centralis (in the hawk two), the elements in that part all contain yellow spherules.

The retina of reptiles closely resembles that of birds. In lizards, according to Leydig, two kinds of elements are distinguishable—one of a slender form, and furnished with a deep yellow spherule ; and others of a broader conical shape, whose apex is coloured with a diffuse yellow pigment. Both these elements, however, it would seem, according to Schultze, should be regarded as “cones.” According to H. Müller, the retina of the chameleon contains only elements of one kind, which must also be regarded as cones. In the cones of Anguis fragilis, which have been subjected to osmic acid, and, apparently, according to Müller, in the chameleon, a peculiar differentiation of the contents of the inner segment of the cones is observable, in the appearance of a conical, strongly refractive body, the base of which is directed outwards, whilst the pointed proximal extremity looks towards the membrana limitons externa, though it does not actually reach it.^* These bodies were supposed by Müller to represent cell-nuclei, but M. Schultze suggests that they are refracting lenses.

Throughout the amphibia a great uniformity exists in the retinal elements. Amongst numerous colossal “rods “are lodged a few very minute “cones,” each of which contains a minute-coloured or colourless spherule.

M. Schultze confirms Henle’s discovery of the presence of one or more transverse lines in the outer granules, or rather on those of the outer granules which are connected with the “rods,” as they are not found on those belonging to the “cones.” These markings appear to be absent in all other vertebrates.

A very full account of the structure and relations of the black pigmentary layer is given, and reasons shown for its being regarded as an element, not of the choroid, but of the retina itself. It consists essentially of a layer of cells containing black pigment, and which send down fine filamentary processes, like the pile of velvet, to fill up the spaces between the outer segments of the “rods “and “cones.”

The paper then proceeds to give an account of the arrangement, &c., of the “cones,” which alone constitute the percipient stratum in the macula lútea. It is shown that as the border of this spot is approached the number of “rods,” in proportion to. that of the “cones,” gradually and regularly diminishes, until at last the former cease altogether, whilst at the same time the “cones “themselves become longer and slenderer up to the centre of the macula ; the direction, also, of the cone-fibres becoming more and more oblique as they radiate, as it were, from the centre of the macula. As is now well known, the layer of “cones “is continuous over the so termed fovea centralis. Some very interesting observations are given on the subject of the relation of the diameter of the “rods “and “cones “to the acuteness of vision, &c. ; and the probability is shown that at the point of junction of the outer and inner segments of the “rods “and “cones,” which differ so much in their refractive properties, and between which, as pointed out by Krause, even in the perfectly fresh state so sharp a line of demarcation exists, the light passing through the retina to the “rods “suffers reflexion upon the end of the inner segment, or upon true percipient nervous point, as it may be termed.

The third section treats of the development of the retina, and especially of the “rods “and “cones,” and it contains many extremely interesting original observations. The author’s study seems to have been principally directed to the development of the eye in the chick. He shows that the pigment-layer of the retina, or the inner layer of the choroid, as some deem it, is formed in the outer coat of the primitive eye-bulb-sac, and that the outer and at first perfectly even surface of the inner coat of the bulb is in close contact with the outer. The surface of the inner fold of the primitive bulbsac is formed by, or rather represents, the future membrana limitons externa. The first indication of the formation of the “rods “and “cones “is visible on the previously perfectly even surface of this membrane in the appearance, about the tenth day of incubation, upon it of minute hemispherical elevations, which are, in fact, the rudiments of those elements into which the elevations gradually grow.

In mammalia the necessary continuous observation is not so readily made, but sufficient has been ascertained to show that the development of the retina in them proceeds in the same way as in the fowl. In fresh embryo calves, in specimens from fifteen to twenty-five centimeters in length, the membrana limitons externa was in close contact with the pigment-layer, and no trace of either “rods “or “cones “was visible. In specimens fifteen to twenty centimeters long, hardened by immersion in “Miiller’s fluid,” or in a weak solution of nitric acid, although the nerve-fibre-layer of the retina was distinct enough, none of the other layers were as yet differentiated from the general substance composed of spindle-shaped cells having elongated nuclei and processes passing to the outer and inner membrana limitons.

In embryo sheep, at the time of birth or very nearly so, “rods “and “cones “were present, but not at an earlier period. They were, however, shorter, and, above all, much more delicate, than in the full-grown animal.

It would appear that in the sheep and other mammals the “rods “are not developed until the differentiation of the other parts of the retina has advanced some way, nor before the end of embryonal life ; but in some instances, as in the rabbit and cat, this is seen in a far more striking manner. Neither of these animals at birth present any trace of “rods “and “cones.” The blindness, therefore, of the new-born rabbit and kitten does not depend solely upon the closure of the lids, but is also associated with an undeveloped state of the retina itself. The “rods “and “cones “do not appear to be fully developed before the thirteenth day, when they are in the same condition as in the calf or lamb at birth.

The development of the “rods “and “cones “in man appears to follow the same course as in the ruminants above named.

The fourth section relates to the differences between the “rods “and “cones,” with respect more especially to their functions. And in it is given a recapitulation of the principal anatomical facts upon which the physiological conclusions or suppositions are based, in the following words :

The author then enters upon the question of the physiological relations of the “rods” and “cones and the following may be taken as a very brief summary of his highly interesting observations on this point.

The organization of the “yellow spot,” and of the fovea centralis, in the human retina, clearly proves that the cones alone are not only sufficient for vision, but also that they possess certain physiological advantages over the “rods.” But it is, at the same time, obvious that the “rods” alone suffice for the purpose of vision, since the bat and other mammals are wholly unprovided with “cones.” But these mammals without cones in the retina prefer the dusk or night to daylight. The question, consequently, arises, what impression communicated through the retina in the dusk is useless ?—by the solution of which we may be guided to some conclusion with regard to the peculiar function of the “cones.”

The visual sense comprises three fundamental impressions, whichhavebeen termed by Aubert “Lichtsinn, “Farbensinn,” and “Raumsinn ;” that is to say, “light-sense,” “colour-sense,” and “space-sense.” It as at once obvious that the light-sense, or the power simply of perceiving luminosity, including [perhaps] quantitative differences in the degree of light, is a fundamental requirement in any, even the simplest, visual organ. For this purpose, it is clear that a single termination of a nerve, or, in other words, in the case of the retina of the higher animals, a single rod or cone, would suffice. And it may also be admitted that a number of such visual points, associated so as to form a single percipient organ, would, in addition to the simple perception of light, also give the power of estimating space, and consequently of conveying ideas of form. These two faculties of the perception of light and of space as conveyed by light are inherent in the eyes of all vertebrates. The “coneless” retina of the bat, hedgehog, and mole, does not, in this respect, differ from the “rodless” retina of snakes and lizards, seeing that the “cones” are, at any rate, quite as fully percipient of light as the “rods,” inasmuch as they equally represent the termination of sentient nerves. It may be assumed that the mere sense of luminosity is more strongly developed in nocturnal animals, as the bat, than it is in the sunshine-loving snake ; so that the former would find a sufficiency of light when the latter was in darkness. This would seem to indicate that the “rods” were more adapted for the simple perception of light than the “cones.”

We have next to consider the colour-sense ; that is to say, the sense by which qualitative differences in light are perceived. To judge from our own experience, which in such a question can be the only guide, the simplest trials will show that, as dusk and darkness approach, the power of perceiving colours ceases at a comparatively early stage. In the evening, though we may see objects well enough, we are quite uncertain as to their absolute or relative colour; We may suppose, therefore, that an animal which pursues its prey only at night, and which habitually frequents dark or obscure places, has no sense of colour, or, at any rate, only needs to distinguish different degrees of brightness in the different colours, as is the case with ourselves in the dusk [or even, in the case of colour-blindness, sometimes even in the daytime]. If we assume, as from the theory of Young- and. Helmholtz we are compelled to do, that the sense of colour resides in a determinate anatomical substratum, we,are justified in concluding that that particular substratum is wanting in the retina of nocturnal animals. The conclusion naturally follows, that the “cones” may, in all probability, be the terminal nerve-organs of the colour-sense.

It should be borne in mind, however, that the “cones” cannot be regarded as exclusively confined to the perception of colour. The colour-sense necessarily includes the light-sense, or is, as it were, superadded to it ; and thus we may conclude that, where the colour-percipient cones are sufficiently closely aggregated, they may also suffice for the sense of space, and thus fulfil all the functions of a retina by themselves alone. The only question, therefore, as M. Schultze remarks, that can arise, is as to whether it is probable that the “cones,” together with the power of conveying impressions of luminosity and space, have not in addition that of conveying impressions of colour, and whether we have any reason, in like manner, to suppose that the “rods” have no such power.

The author then proceeds to show, in reference to the experiments of Purkinje, Hueck, Helmholtz, Aubert, and Schelske, that, although the sense of colour exists throughout the human retina, it is most acute in proportion to the preponderance or number of the “cones “over the “rods,” and that the latter alone are unable to convey impressions of colour. He also points out that the probabilities that this function resides in the “cones” is strengthened by the fibrillated structure of the “cones “and their filaments, which is in accordance with the well-known theory of colour-perception, propounded by Young and Helmholtz, that at least three different kinds of fibre must be required for this perception. Each “cone,” therefore, in the mammalia and fishes, having this compound structure and all being alike, it would appear to follow that all are equally capable of perceiving every variety of colour. And his argument is still furtherstrengthened by the consideration that, inasmuch as all or nearly all the “cones “in a bird’s retina are furnished with a coloured spherule, through which all the light reaching the percipient part must pass, it would be absurd to suppose that they were incapable of receiving impressions of colour, for which, so far as shown by that circumstance, they alone would seem to be fitted. Furthermore, it is to be borne in mind that all the “cones “in a bird’s eye do not contain spherules of the same colour, and that some are without any, whence we may conclude that in all probability the differently coloured “cones “are adapted for the perception of mono-chromatic light corresponding to that of the spherule contained in them, and that each is not, as in the mammal, capable of conveying equally impressions of all colours. And this view is curiously in accordance with the circumstance that the “cone ‘‘-filaments in the bird are scarcely thicker than those of the “rods.” Whether this be the’case with the filaments proceeding from the colourless” cones,” has not been made out. But it may be that these “cones” are adapted for the reception only of the violet rays, which would, of course, be absorbed in their passage through the coloured “cones.”

The structure of the owl’s retina, in contrast with that of diurnal birds, may be cited in support of the same argument. And the author refers to a suggestion of his own, made in a former paper on the macula lútea,^* that the intervention of the yellow spherule in birds, and of the yellow colour in the human macula, may serve for the interruption of the more powerful photo-chemical rays in their passage to the delicate percipient tissue. ^†

This part of the paper concludes with a highly interesting disquisition respecting several other points connected with the simple visual sense and the estimation of sizes and forms, &c., for which the reader must consult the original.

In his researches on the retina M. Schultze has found the greatest advantage in the use of a solution containing 1 to - of osmic acid (OsO₄); and he recommends that a solution of that substance containing 1 per cent, should be kept at hand, which can be diluted at pleasure. Microscopic preparations made with it he prefers to keep simply in water.

The black colour which is assumed by the preparation, even within a few minutes of its immersion, is at first uniform throughout. But subsequently the different parts of the retina exhibit slight differences, the optic nerve-fibres and the molecular and intergranular layers exhibiting the deepest tint. In frogs and fishes the deepest colour is seen in the outer segments of the “rods.” In this way may be obtained preparations in which the outer segment is of a deep black colour, whilst the inner is almost uncoloured, the line of demarcation between the two being very abruptly defined. A similar difference is observable also in mammals, but not so constantly, and under circumstances which cannot at present be explained. But the demarcation between the segments is always well defined, and the author can recommend no better medium for the examination of the “rods “and “cones “thanjosmic acid. A special advantage of the osmic-acid solution is that it hardens the elements of the connective-tissue framework more slowly than the nervous; and another is that, except in very strong solutions, it does not produce granular coagulation either within or without the elementary parts of the retina.

The observer is cautioned against the injurious effects of osmic acid upon himself, unless great care be taken.

Another medium greatly employed by him is what he terms “lod-serum,” or iodized serum, which is used for the immersion of fresh dissections of the eye and other parts—the most delicate tissues, such as the retina, remaining for a long time unaltered in it. It is prepared from the amniotic fluid of the calf, to which a sufficient quantity of tincture of iodine is added to give it a faint yellow colour. And he has found that an albuminous fluid of this kind may be kept unaltered for any length of tune if a very minute quantity of bromine be added to it. But as bromine acts very powerfully in causing cells, &c., to contract, the quantity added to the iodized serum must be less than will remove the whole of the yellow tint.

[It is not improbable that a few drops of carbolic acid would answer the same purpose as the bromine, and perhaps the iodine also.]

* From an inspection of Mr. Lewis’s drawings we feel bound to say that the perforations appear to us more frequently hexagonal than pentagonal.— ED.

† Vol. i, p. 188.

* Supposing the perforations to be six-sided, as they appear to us, and not five-sided, as Mr. Lewis thinks they are, an explanation is not so difficult. The shape of the spark itself is, in all probability, that of a more or less regular cylinder, whose section is a circle. The resistance offered to the passage of the spark by the perforated paper or card, acts at the point of passage on its cylindrical form in a manner analogous to that in which equal pressure from all sides acts on a solid cylinder, rendering it hexagonal; though the spark must not be regarded as anything but a condition of the atmosphere. The cases of the basaltic pillars of the Giant’s Causeway and of agglomerated soap-bubbles are well-known instances of this law of pressure.—ED.

* This is probably the “albuminous substance which, in chromic-acid preparations, retires as an opaque granular mass towards the outer end of the body of the cones,” noticed by Mr. Hulke Proc. Roy. Soc.,’ xiii, p. 109).

“With the enlargement of our knowledge of the structure and disposition of the two elements composing the percipient layer of the retina—the ¹ rods ‘and ‘cones’—arises the question whether we are thus in a condition to attempt the problem of the hitherto unknown physiological distinction between them. We hear that, at any rate, the direction in which the solution of this question is to be sought may now be indicated with some degree of certainty, and I will end eavour briefly to state my views, as follows:

“The anatomical facts upon which we have to rely, shortly recapitulated, are these :

“1. The difference in size and form. This is manifested more particularly in the so-termed inner segment, which in the ‘rods ‘is always sharply defined from the outer segment, but which may also be distinguished as a separate element also in the ‘cones.’ The inner segments in both the ‘rods ‘and ‘cones,’ in the perfectly fresh condition, consist of an apparently almost structureless substance, but which very rapidly, after death and in all preservative media, coagulates into a more or less distinctly granular matter. This substance, to judge from micro-chemical reactions, most nearly resembles albuminous matter, as, for instance, the protoplasm of young cells. An essential distinction between the substance of the inner segment of the ‘rods ‘and of the ‘cones ‘is manifest in the circumstance that solutions of osmic acid of a certain strength produce in that of the cones a very distinct parallel striation, which, under similar conditions, I am unable to perceive in the inner segments of the ‘rods.’ No universal distinction exists in the absolute diameter of the inner segments, as, for instance, in the human retina; for although the cones throughout by far the greater part of the retina are fully twice as thick as the rods, their inner segments in the fovea centralis are quite as slender as those of the ‘rods.’ The outer segments or shafts consist of a more highly refracting substance, which after death coagulates in a different manner from that composing the bodies. This substance does not become granular, like protoplasm, but either hardens into a homogeneous mass or shrinks and curls up in a peculiar manner, at the same time cracking, generally transversely, but sometimes also longitudinally. That an external tunic and contents—a cortex and central filament—can be distinguished in them I hold to be highly improbable. The outer segments of the ‘rods ‘are cylindrical, though a very slight attenuation towards the choroid may occur (frog) ; on the other hand, the outer segments of the ‘cones ‘are of a decidedly conical form, the apex pointing outwards, and usually terminating below the summits of the rods.

“2. Avery remarkable difference between the ‘rods* and ‘cones’ is presented in.the filaments proceeding from them to the external granule-layer. The filaments belonging to the ‘cones’ are of considerable thickness;which sometimes is as much as 2—5 micro-millimeters; they exhibit here and there a delicate longitudinal striation, as if they were composed of parallel fibrils ; and they always break up on the upper surface of the intergranular layer into an indeterminate number of extremely delicate fibrils, which are lost in that layer.* The fibres proceeding from the rods, on the contrary, have a scarcely measurable thickness, and they can only be traced to the surface of the intergranular layer, where they apparently terminate in a. minute enlargement whose nature is at present obscure: ‘Each filament, whether belonging to a ‘cone ‘or ‘rod,’ is in some part of its course connected with a cell—an outer granule—so that the outer granules may be divided into ‘rod’ and ‘cone-granules,’.of which the latter, at any rate in the mammalia, are the larger. Both kinds of filaments present all the characters of nerve-fibres, and much resemble those of the optic nerve-layer, and, on the other hand, they are manifestly distinguishable from those of the trabecular framework.

“3. At the yellow spot of the human and simian retina ‘cones ‘only exist. Close to its periphery, however, ‘rods ‘become interposed between them, and at a few millimeters from the middle of the spot they are present in the number of two to three between each two ‘cones,’ a proportion which is continued uninterruptedly up to the ora serrata. In proportion as they become crowded together at the macula lútea, their fibres, as well as those of the ‘rods’ interspersed among them, assume an oblique direction, radiating, as it were,

* In a valuable communication to the Royal Society, read in June, 1866, on the “Chameleon’s Retina,” Mr. Hulke states “that from the inner ends of the cones fine fibres proceed obliquely from the outer to the inner surface of the retina in a radial direction from the centre of the fovea to the periphery of the retina.” These fibres connect the cones with the cells of the outer granule-layer ; they next form a thick plexus at the inner surface of this layer, which he terms the “cone-fibre-plexus then traverse the inner granule-layer, in which they connect themselves with round and roundly oval cells, and are continued through the medium of the ganglion-cell-like cells of this layer into the granular (molecular layer, Schultze), where they join the processes directed outwards from the cells of the ganglionic layer. “Thus,” he says, “they constitute an anatomical path between the cones and optic nerve-fibres.”

This, if confirmed by future observation, is a most important fact, and one of great import with relation to the apparently more direct and immediate communication between the “cones” and optic nerve-fibres than would seem to obtain with respect to the “rods.” from the centre of the macula in a meridional and forward direction, in order, after a longer or shorter course, to reach the outer granular layer.

“4. In most mammalia the relative number of ‘rods ‘and ¹ cones ‘is exactly the same as in man, with the exception, of course, of the macula lútea. But in many the cones are altogether absent. This is the case in animals which prefer darkness to light, such as the bat, hedgehog, mole, mouse, and probably a great many others. In the rabbit, which, as is well known, in the wild state inhabits subterranean passages, there are, it is true, indications of cones, though these appear to be in quite a rudimentary state.* The cat has distinct though slendeicones, which are placed wide apart, so that room is left between them for twice or thrice the number of ‘rods ‘than in the human retina.

“5. Birds have many more ‘cones ‘than ‘rods,’ the former, in fact, standing to the latter in the inverse proportion to that in which they occur in the human subject. In both the foveœ centrales of the falcon ‘cones ‘only exist [as well as in the single fovea centralis in some other birds). But the owls almost resemble the bat, their retina containing but very few cones and an enormous proportionate number of rods. In their retina scattered ‘cones ‘only occur at wide intervals, and these are so overcrowded by the very long outer segments of the ‘rods ‘as to be seen with great difficulty.

“6. The ‘cones ‘in birds are distinguished by a very remarkable character. The greater number of them are furnished, at the end of the inner segment and immediately in front of the point of attachment of the outer segment, with a highly refractive globule, for the most part of a deep yellower red colour-, anything analogous to which, so far as is at present known, is wanting in all mammals. The yellow globules are more numerous than the red. The coloured globules have a diameter precisely corresponding with that of the base of the outer segment, so that no light can reach that part without passing through the globule. The few ‘cones ‘which have no coloured globule contain at the corresponding point a strongly refractive colourless body, apparently of the same kind. The few ‘cones ‘existing in the owl’s retina are furnished with pale yellow or colourless globules. Red globules are entirely wanting in the retina of those birds (Strix aluco, noctua, aniflammea).

“7. Among reptiles, in some, as the turtle, the retina appears to present the same structure as that of birds. Lizards and snakes have only cones, and in some instances these contain pigment-globules in the same situation as in birds {Lacerta, sp. Anguis fragilis), whilst others are without these coloured elements (chameleon and snakes).

“8. The amphibia (frog, toad, triton, and salamander) have very thick rods and very minute cones, but in each of the latter is a bright yellow or colourless globule situated between the outer and inner segment.

“9. The osseous fishes, so far as researches have hitherto gone, appear to possess rods and cones like the mammalia ; and the latter are without coloured globules. Cartilaginous fish, on the other hand, as the ray and dog-fish, are wholly without ¹ cones,’ like the bat among mammalia.

“10. The difference which in mammals and fish is so apparent