were known to have been taken under unfavourable conditions of illumination, the results appear to me to be quite trustworthy, as they conform when grouped in the usual manner to the wellknown "law of error." The groups increase in a fairly uniform manner from four observations at and under 5 feet, up to 272 observations at 50-60 feet, and then diminish in a similar manner to four observations at 110 feet and upwards; the mean or largest group being, at 57.5 feet, identical with the theoretical distance at which the dots are visible to persons with good eyesight. The difference, moreover, between the eyesight of town folk and of country folk, as shown by these statistics, is barely 4 per cent. in favour of the latter, a difference which is probably to be accounted for by the better light in the country. These results would seem to show that the varying distances between the dots on the test card is not a vital objection to their use, as the result depends on the power of the eyes to separate the dots which are nearest together, and these are always separated by one diameter. It would be desirable, however, in altering these tests, or in devising new ones, to avoid this possible source of error and arrange the test objects, whatever form they may take, at equal distances from each other. [A diagram was here exhibited showing the effect of grouping the test dots at distances of one, two, three, four, and five diameters apart, and the increasing distances at which each set is visible.]1

In arranging tests for eyesight sufficient attention has not been given to the interference with their use which astigmatism produces. This defect of sight is much more common than is generally supposed, being according to my experience rarely absent in adult persons, but has not received so much attention as other defects of sight because it does not in some of its forms materially interfere with the ordinary use of the eyes. In these cases there is a meridian of the eye which possesses the proper focal length lying between two portions which are imperfect. In the ordinary use of the eyes the meridian of good sight is made, by the rapid movements of the eyeball, to range over the object to which attention is directed and a sufficiently clear image of it is formed on the retina, but it is quite otherwise when the attention is fixed on a small object like a test dot or a testtype. In this case the test object becomes elongated in a direction at right angles to the meridian of good vision, and hence dots

1 It must be borne in mind that these test dots are used in the British army as a minimum test, and any recruit who cannot distinguish them at a distance of 15 feet is rejected as unfit for military duties, as he would be unable to see a bull's-eye target 2 feet square at a distance of 600 yards. The instructions for carrying out this examination of recruits are unfortunately introduced in the "Notes and Queries," and they must be very puzzling to persons not accustomed to testing eyesight.

separated by only one diameter are seen together and assume the appearance of a continuous line before they disappear from sight by increasing distance. This common form of astigmatism is attributed to irregularities of the surface of the cornea, but there are other forms due to irregularities in the form of the lens which have received little attention from ophthalmologists, and which are often confounded with myopia, with which they have many features in common.

To eliminate the interference of corneal astigmatism with the use of small test objects, and as the direction of the meridian of good sight is not constant but varies in different persons, and possibly in different races, I have proposed the adoption of a series of concentric circles, which, as they embrace the whole field of vision, must be visible to astigmatic eyes in one direction, and in this direction therefore the lines may be counted. (See figure.) These circular diagrams are a most delicate test for all forms of astigmatism, apart from other forms of defective eyesight.

There are many reasons why we should hesitate to accept the statements often made relative to superiority of the eyesight of savages over civilised peoples till the sight of both has been submitted to some rigid test similar to those I have referred to. I have myself been much among savages (several tribes of North American Indians, Australians, and the Hill tribes of India) without noticing any remarkable manifestations of good eyesight apart from that which was due to a special knowledge of the objects observed. The travellers' tales on this subject are to be accounted for by faulty standards of comparison. Many-probably mosttravellers are ignorant of the quality of their own eyesight, and

VOL. XV.

K

sailors, from whom many of the stories reach us, are a wonderloving class, and not always the masters of their imaginations. The cases, moreover, which have been recorded are only those of persons who possess remarkable eyesight, and not the average of a large number of the same race living under similar conditions of life, and we do not know the best eyesight among civilised people with which to compare these cases. Eyesight equal to a visual angle of one minute has been accepted as the average of civilised persons, but this necessarily represents only half the acuteness of vision which some persons possess. InspectorGeneral Lawson, recording his own experiments in the Report of the Anthropometric Committee for 1881, tells us that he could at one time distinguish a flagstaff at Aldershot, the smaller diameter of which was 6 inches, at a distance of three miles, under favourable conditions of the atmosphere, the visual angle subtended being only 6.7 seconds, or a ninth part of the commonly received visual angle of one minute; and he further states that it could be seen by other persons under similar conditions. Among the observations collected by means of the army test dots already referred to, four men are returned as seeing them at 110 feet and upwards, that is to say, at about double the average distance. It is obvious, therefore, that if the test is to be one of who can see objects at the greatest distance, we must begin by ascertaining the best eyesight to be found among ourselves. This would be, however, a very difficult and unscientific method, and it is only by obtaining the averages of a large number of observations among savages and among civilised races, and comparing them together, that any useful results can be obtained.

DISCUSSION.

The PRESIDENT remarked, before the first paper was read, that the question of the relative keenness of sight of savage and civilised races had lately been brought prominently into notice by Mr. Brudenell Carter. It concerned a matter of fact, and was one which this Institute might legitimately undertake to get solved. The question was not as to the greater quickness of observation and of perception of the savage, because on that point all were agreed ; but whether his eye, as an optical instrument, was superior to that of a civilised man. To solve this question, satisfactory tests had to be thought out equally suitable for use in savage and civilised countries, and if this could be done there was little doubt that this Institute was capable of inducing many travellers to apply them.

After the reading of the papers, the PRESIDENT expressed his satisfaction at finding that there was a concurrence of opinion on one essential point, namely, on the superior merits (under clearly explained conditions) of test dots or circles. He himself thought

that for the purpose of travellers the number of those dots should be limited to one or two. He exhibited a thin octagon zinc plate, 5 inches in width, with holes in it to which paper test circles, of whatever construction might be adopted, could be attached by threads. The plate would be propped on one or other of its eight sides, giving that number of varying test positions.

ANTHROPOLOGICAL MISCELLANEA.

ETHNOLOGICAL NOTES on the ARABS of ARABIA PETREA and WADY

ARABAH.

By E. GORDON HULL, M.D., B.A.

THE principal tribe inhabiting the Sinaitic peninsula, south of the Hadj road from Suez to Akabah, is the Towara, and amongst them I spent some weeks in the winter, 1883-4, while assisting my father, Professor Hull, to make a geological survey of the country for the Palestine Exploration Fund; the other tribe with which we came in contact being the Alawîn, who live in a territory bounded on the north by Wâdy Mûsa and Petra, on the west by Wâdy Arabah and Gulf of Akabah, and which extends somewhat indefinitely east and south. Their Chief, Muhammed Ebn Jad, is subsidized to protect the Egytian Hadj.

We also encountered the clan which inhabits the Wâdy Mûsa, who appear to be of quite a different race. Their skin is very much fairer than that of the Arabs, they are considerably taller, and most of them have aquiline noses-in fact they look rather like Jews. The Arabs look down on them, because they cultivate the land, calling them "Fellaheen," and it is rather a problem as to where they came from. With regard to the Towara, it may be said that they are a small race, but their limbs and bodies are well formed and proportioned, and they are capable of enduring great fatigue on a diet consisting principally of boiled rice and butter, with unleavened bread, coffee, dates, and water.

I measured twenty-six adult males, picked men of the tribe; taking three measurements, that is, height, chest round nipple, and length of right arm from acromion to tip of middle finger. The average height was 5 feet 4 inches; average chest-measurement, 31 inches; average length of right arm, 27.6 inches. The maximum chest-measurement in a mau of 5 feet 11 inches was only 344 inches, and the minimum in one of 5 feet was 30 inches. Yet with such insignificant chests they were splendid pedestrians and mountaineers, and did their day's march without a murmur. Certainly, they were all in very good condition, for I do not suppose there was an ounce of fat among the whole tribe.

The following are the detailed measurements. As they had an objection to being quite stripped, I put the tape round their chests over their thin calico shirt, so that the measurements are slightly in excess, 25 to 5 of an inch.

The chest-measurements were taken at the level of the nipples. with the arms down by the side, at the end of expiration.