must be added to the German cranial index to make it correspond with the French cephalic index; but owing to his special methods Welcker's indices average, as Ripley' points out, two units below other German indices.

1

About the same time that Ripley had come to the foregoing conclusion, Dr. Ammon 'published a paper on the same subject, as the result of a very careful inquiry. As this is an important matter, I give Ammon's conclusions for the convenience of those who may wish to correlate the cranial measurements made by the two chief Continental schools.

"Length. To obtain the French length from German measurements add 1 mm. to the brachycephals (80-84.9) and hyperbrachycephals (85-89.9); 1.1 mm. to the mesaticephals (75-79.9) and ultra-brachycephals (90-94.9); 1.5 mm. to the dolichocephals (70-74.9), and 1.4 mm. to the extreme brachycephals (95-100).

66

Breadth.-Identical in both methods.

Index. The French index is obtained by deducting half a unit from the German index; or, to be more exact, deduct .6 from the dolichocephals and .7 from the extreme brachycephals instead of .5.

"Inversely, to transform the numbers obtained according to the French method into those of the German method, one must deduct 1 mm. from the length, etc., and add .5 to the index."

I

3

Collignon points out that one is too apt to attach to the terms dolichocephaly and brachycephaly a concrete sense which they should not have.

Thus a human variety may by the proportions of the body and its members and by the characters of the face and

1 Loc. cit., p. 520.

2 Otto Ammon, "La Corrélation entre l'indice céphalométrique de Broca et celui d'Ihéring," L'Anthropologie, vii., 1896, p. 676.

Mém. Soc. d'Anth. de Paris (3), i., 1895, p. 23.

the length of the skull, manifestly resemble a dolichocephalic race and yet have a cephalic index of 81 or 82. Or another variety may have a long and broad body, short legs, a low flat face and short head, and other characters of a brachycephalic people, and yet have a narrow head, giving an index of 78 or 79. Collignon would respectively allocate these aberrant varieties to the dolichocephals and to the brachycephals respectively.

One knows that in a race as pure as possible there may be a range in the cephalic index of 13 units, that is to say, a pure race having a mean index of 84 may normally vary between 78 and 90, and, inversely, a race having a mean index of 77 may oscillate between 71 and 83.

There are thus normally, and without any foreign influences, some individuals of a brachycephalic race who may be relatively dolichocephals, and, conversely, there may be brachycephalic individuals in a dolichocephalic race.

When the mean anterio-posterior diameter of the cranium exceeds 190 mm. in the living, the head should be considered as long, and the race to which it belongs is ranged as a whole in the great dolichocephalic group, rather than in the group of brachycephalic races. It will be noted that Collignon for the sake of brevity of description ignores a mesaticephalic group.

Collignon gives' the following examples:

On page 77 of the same Memoir he suggests the index of 82 for the limit of dolichocephaly and brachycephaly in the living subject.

As these practical instructions are intended for ethnographical survey work in the field and not for craniological purposes, I will not give instructions for skull measurements, but will refer the reader to such works as Flower's Osteological Catalogue (Part I., "Man") of the Royal College of Surgeons (1879); Garson, in Notes and Queries on Anthropology (1892); Topinard's Eléments d'Anthropologie génerale (1885); Broca's Instructions crâniometriques (1876); De Quatrefages et Hamy's Crania Ethnica (1882); Garson, “The Frankfort Craniometric Agreement, with Critical Remarks thereon," Four. Anth. Inst., xiv., 1884, p. 64; Schmidt's Anthropologische Methoden (Leipzig, 1888), and Macalister's Anatomy; Quain's Anatomy, etc. Professor Macalister is writing a book which will thoroughly cover this ground.

The only skull measurements I have referred to are the length, breadth, and height, and the two measurements required for the nasal index; the latter have been sufficiently described on page 95. The length of the skull is taken by the French, English, and American anthropologists from the glabella (the middle line between the brow-ridges) to the greatest diameter obtainable behind in the middle line; this gives the greatest obtainable mesial length. All anthropologists take the greatest breadth wherever found above the levels of the ear-holes; care must be taken that the callipers are held at right angles to the median longitudinal

(sagittal) line, and that the two ends of the callipers are at the same level. The height of the skull is the diameter between the basion (the middle point of the anterior edge of the foramen magnum) and the bregma (the point on the vertex where the frontal and interparietal sutures meet).

Nasal Measurements.-Dr. R. Collignon, who has paid more attention than any other anthropologist to the nasal index in the living, draws attention' to the great care that is required in taking the measurements, as, owing to the low units of the dimensions, even small differences in the method employed will lead to considerable differences in the results. The nasal index is an extremely delicate measurement, and it is only on account of its exceptional ethnological importance that it claims so much attention and trouble.

We have already seen that the index is the ratio of the maximum breadth of the base of the nose, that is, of the wings of the nose, to its height, the latter being taken from the root to the point of insertion of the septum.

There is no difficulty about the first measurement, except that care must be taken not to squeeze the wings and to guard against laughter, which often happens, and which gives a false platyrhiny by the distension of the nostrils. Distension is also caused by breathing through the nostrils; this can be stopped by the subject keeping his mouth open.

The second measurement is less easy, and in certain cases is of extreme difficulty owing to an uncertainty in selecting the upper point of measurement.

Among the great majority of Europeans the condition does not exist; the origin of the root of the nose is generally seen at a glance with the greatest ease. It suffices to look at the face in profile, and to place one arm of the compass at the deepest point seen. But among the yellow or black

1 64

La Nomenclature quinaire de l'indice nasal du vivant," Revue d'Anthropologie (3), ii., 1887, p. 8.

races and among some Europeans it is by no means the same; the line of the forehead very often passes on to the nose in a regular curve, which does not present any change of direction. In this case, Topinard recommends that the subject should be looked at full face; usually one can distinguish a slight horizontal fold which is the desired spot; or by stroking vertically from above downwards upon the skin of the lower part of the forehead, the fold is indicated. in the great majority of cases with great clearness. This should be lightly marked with a pencil.

To recapitulate: In order to determine the height of the nose, stand to the right of the subject and look at his profile; if the exact spot where the root begins is clear, place the lower arm of the sliding compass against the angle which the septum makes with the upper lip, without pressing it too much, and gently bring down the upper arm to the upper spot. If there is any doubt about the latter, press down the skin of the forehead with the left hand, always looking at the profile, so as to appreciate the change of direction which is produced, mark it, and proceed as above.

Instead of the ordinary sfiding compass, which may have sharp points, and which is usually difficult to use with precision for these measurements, Collignon for many years has employed a small, light, box-wood sliding compass, which can be obtained from any French shoemaker, and which costs only I fr. 25 c.

It cannot be too strongly insisted upon that the subject must be measured in profile, as a measurement taken in full face is nearly always erroneous.

One further detail must be added: It is necessary to hold the compass in the right hand, to apply first the lower arm against the septum and to hold it there, and gently to make the other arm descend to the upper point of measure