The test-types in common use are so proportioned that, at some stated distance, they subtend a visual angle of five minutes in height, and an angle of one minute across their limbs or parts; and persons who can read them easily and correctly under these conditions are said to possess normal vision. This very moderate standard is constantly exceeded.

Lord Rayleigh has lately introduced a new element into the question, by suggesting that the defining power of the eye, as an optical instrument, is limited by its aperture and by the wavelength of light. He has published a rough calculation on which he founds the conclusion that the limit of sight would be reached under a visual angle of "about" two minutes. He stated that this limit had been "approached" by civilised physicists, and hence inferred that there was little room for uncivilised men to surpass them. It would be rash for me to assert that Lord Rayleigh is wrong in his principle, but he is obviously wrong in his application; for the limit which he assigns to vision is, as I have already stated, one which civilised physicists have not merely approached, but very far exceeded. There is, therefore, at the very least, a serious error in his calculation; and I may had that there is no such variation of visual acuteness, in relation to changes in the diameter of the pupil, as his hypothesis would seem to require. Without denying that there is an optical limit to the defining power of the eye, I am yet disposed to believe that this limit has no bearing upon the question at issue, and that the practical limit of visibility depends, in healthy and well-formed eyes, entirely upon the anatomical formation and the functional sensibility of the retina.1

The next point to which I must call attention is that the conditions shown in the diagram are not of universal occurrence, and that the actual relation between the magnitude of the retinal image and the magnitude of the visual angle is by no means of a simple character. The dioptric system of the eye is complicated, and one effect of its complexity is that the point of divergence of the axial rays within the eye is not identical with their point of meeting, but is posterior to it. Moreover, the distance between the meeting-point and the point of divergence differs in different circumstances. A large proportion of eyes are so constructed that the focal length of their refracting media is identical with the length of the antero-posterior axis.

After this paper was written, the error in Lord Rayleigh's calculation was pointed out in a letter to "Nature," and was acknowledged by his lordship. The optical limit of vision would only be reached under a visual angle of 28.9 seconds; and hence, as the finest seeing hitherto accomplished by a "civilised physicist" has been under an angle of 50 seconds, this best recorded performance represents scarcely more than one-half of the acuteness which would be theoretically attainable.

Persons who have such eyes are said to be emmetropic, that is, their eyes are in correct measure or proportion, and hence they possess the largest attainable range of functional activity. Persons whose eyes are not in this correct proportion are said to be ametropic, or out of measure; and ametropia presents two chief varieties, hypermetropia, in which the focal length is greater than the length of the antero-posterior axis, and hypometropia, commonly called myopia, or short-sight, in which the focal length is less than the length of the antero-posterior axis. Generally speaking, it may be said that hypermetropia, or flat eye, is a state of arrested development or of degeneration, and that myopia is a state of malformation, often originally inherited in some degree, but always aggravated by use and often complicated by disease. Both forms of ametropia are liable to be further complicated by astigmatism, a state in which the ametropia differs in degree in different meridians. In every variety of ametropia the cardinal points of refraction are disturbed, and different relations are produced between the visual angle and the divergence of rays within the eye. A similar disturbance is produced by the adjustment of the eye for near vision; and, moreover, in the ametropic, unless the defect is corrected by optical means, the result of the divergence of the rays is to produce upon the retina only a dispersion circle, or smudge, instead of a defined image. Before we can be sure, when comparing visual angle with visual angle, that we are also comparing retinal images of similar magnitude and clearness, we must completely correct ametropia by lenses, and must place our test objects at such a distance from the eye as to exclude adjustment for near vision. The smallest distance

at which this can be done is 20 feet, within which range the divergence of the rays proceeding from a point becomes appreciable; and hence all so-called vision-testing, which has been conducted by objects placed less than 20 feet from the subjects, and without the careful correction of ametropia, may be dismissed from consideration, on the ground that there is no common measure for any results which may have been obtained. The external conditions having been the same, the conditions within the eyes would be too various to furnish materials for the judgment. To compare the vision of an emmetrope with that of an uncorrected ametrope, especially after testing them at a distance which calls some ocular adjustment into play, is to compare things between which there is no possible common

measure.

In testing the vision of uncivilised men, it would manifestly be impossible for travellers, generally speaking, to undertake the correction of ametropia, which must therefore be excluded, as

far as possible, by selecting as subjects those persons whose vision is of the average quality of the race. Test-types must be rejected, not only because the subjects would be unfamiliar with the names and aspects of the letters, but also because these are not scientifically accurate, some of them being more legible than others. Single dots or marks, or dots or marks separated from each other by large intervals, must be rejected, because the power to discern a single dot does not depend upon acuteness of vision, but upon sensitiveness to small variations of light. The true test of acuteness is the power to separate two objects which are seen by the same visual act, and the best objects for this purpose are Burchardt's Internationale Sehproben, which consist of black circles or dots, arranged in groups, in such a manner that each circle is separated from the contiguous ones by intervals equal to its own diameter. The groups are numbered, the number upon each indicating the distance, in metres, or parts of metres, at which each dot subtends a visual angle of 215', equal to 1′ 56′′ English. Nothing larger would be required than the group the members of which subtend this angle at 6 metres, roughly 20 feet; and the method of proceeding would be to place the subject at a greater distance than this from the dots, and to cause him to approach until he could count the components of each group. The distance at which this could be done should be measured and recorded; and the experiment should be repeated often enough to exclude guessing or accidental sources of error.

Suppose a traveller to have established, in this way, that the acuteness of central vision, in an uncivilised man, was no greater than that which is often found among ourselves, he would still be only on the threshold of the inquiry. There would yet be three possible conditions which might confer upon the uncivilised man a far greater degree of real acuteness than the testing would suggest. He might have greater sensitiveness to colour, greater sensitiveness to light, and acute vision over a larger retinal area. Each of these advantages would confer upon him a much increased power of readily perceiving small and distant objects, and of discoving their nature.

The percipient elements of the retina, of which I have already spoken, are of two kinds, rods and cones, of which the latter are the more highly developed and the more sensitive. In the human retina, as far as it has been examined, there is a small central area, the seat of the most acute vision, in which the elements are all cones. In a zone immediately surrounding this area, each cone is surrounded by a single circle of rods. In portions of the retina still more peripheral, each cone is surrounded by a threefold or fourfold circle of rods. In the

retinæ of birds, the cones are much more abundant in proportion to the rods; as abundant, in the most peripheral parts, as they are in man in the immediate vicinity of the centre. Not only so, but the cones are of more specialised construction than in man, and each contains a coloured globule, evidently an organ which ministers to a highly developed colour-sense. In bats, on the other hand, there are no cones at all, and the retina is furnished only with rods. It is probable that analogous variations of structure, the results of modes of use extending over many generations, may occur in different families of mankind.

In order to test acuteness of colour-vision, I would suggest that travellers should be furnished with cards on the principle of Burchardt's tests, on which groups of dots should be painted in colours of very slight intensity, and that the distances at which the dots composing these groups could be counted should be recorded. The selection of the colours for this purpose, and the determination of the degree of saturation, would be matters requiring careful and detailed consideration.

În order to determine degrees of sensitiveness to light, I have had an instrument constructed, though only in a rough manner, for exhibition this evening. It consists of a tube, fitted to the orbit at one end, so as to preclude the entrance of light, and closed at the other end by a plug, upon which may be fixed a card, marked with lines, dots, or other figures. At one side of the tube is a shutter to admit light, which is diffused by passing through ground glass. The shutter should be slowly opened until the objects become visible, and the area of the opening can then be read off by graduations at the side. The objects used should include colour dots as well as black ones.

With regard to the last point, the area of retinal sensitiveness, we have abundant material for comparison. In civilised man, as a rule, the sensitiveness of the retina diminishes with much regularity as we leave the centre. For a distance of from onefourth to one-third of a degree from the centre, it retains the central acuteness. At a distance of half a degree it is reduced from about four-fifths to about two-thirds of this acuteness. Up to an angle of 40 or 45 degrees, the reduction in the acuteness is progressive, and is expressed by a fraction which has unity for its numerator, and for its denominator a number ranging from twice to 4 times the number of degrees from the centre. Thus, at 15 degrees from the centre, the acuteness would be less than that of the centre in the ratio of from 30 to 67 to 1; and in farther distances the decrease is still more rapid. On the horizontal meridian, in the outer part of the field, the diminution of acuteness is less than elsewhere; and it is comparatively large on this meridian between the second and

the twelfth degree, that is to say, between the centre of the retina and the entrance of the optic nerve, but these differences may be disregarded for practical purposes. Lateral vision is tested by instruments called perimeters, some of which are very complicated, but I have had a simple perimeter constructed for the purpose of the investigation which we are now discussing. If the instrument is held by the angle, and the piece of wood is pressed against the chin, one eye being closed, any selected object of vision may be placed in the slide, and may be carried laterally, while the open eye is kept fixed upon the central point. Either black dots or coloured dots, or both, should be used as objects; and I should be quite prepared to find that both would be discernible, by the eyes of some savage races, in lateral positions in which they would be quite lost to the eyes of civilised men.

EYESIGHT of SAVAGE and CIVILISED PEOPLE.

BY CHARLES ROBERTS, Esq., F.R.C.S.

THE proposal to test the relative eyesight of savage and civilised races is by no means a new one to English anthropologists. In the small volume of "Notes and Queries on Anthropology" drawn up by a Committee of the British Association, consisting of such well-known members of this Institute as C. Darwin, E. B. Tylor, Col. Lane Fox (Pitt Rivers), Dr. Beddoe, our President, and many others, tests and instructions are given for this purpose. These tests are the set of dots employed by our army surgeons for testing the (minimum) eyesight of recruits, and consist of a series of dots inch square, grouped in a variety of ways to prevent guessing or imposition on the part of the person under examination. These army test dots were largely used by the Anthropometric Committee of the British Association which closed its operations in 1883, but as the results in some instances were not satisfactory they were given up in favour of Snellen's test types. The objections to the use of the army test dots were (a) that some of the dots being placed at unequal distances from each other they were distinguishable at varying distances; and (b) the great distance at which the dots are visible to persons with good eyesight (theoretically 57 feet) in a great measure prevented their use in towns where sufficiently long and welllighted ranges were difficult to find. Since the Anthropometric Committee issued its final Report I have analysed the returns of eyesight sent in, and after eliminating the observations which