Brahmans descended from east-coast Tamil Brahmans, and other classes that I have not copied out; the greatest range is found among the Pariahs of Madras City. Mr. Thurston measured one very dark-skinned Tamil Pariah cooly, who was 5 ft. 3 in. (1608 mm.) in height, and whose nose was 40 mm. in height, 42 mm. in breadth, with an index of 105 (Plate II., Fig. 1).

The least variable, that is, the least mixed, of these groups, have low mesorhine indices (average 75.5); the most variable have higher indices. The Paniyans, with the highest indices of all, have a moderate amount of range.

In order to compare among themselves the Badagas, Todas, Kotas, and Paniyans, it will be necessary to take other data into consideration, so I have selected a few from Mr. Thurston's tables.

The Todas possess exactly the same average stature as the 89,000 Germans, whose measurements are given by Gould, and just miss being included with the English among the very tall races of the world. Between the Todas and the next tallest class measured by Thurston, the Sheik Muhammadans (1645 mm.), there is a well-defined gap of 51 mm. (2 inches). The tallest men he came across were a Toda (1850 mm. = 6 ft. in.) and a Badaga (1832 mm. = 6 ft. 4 in.). The Paniyans have the shortest average, and also have the relatively longest arms.

Measurements which are useful in some other places have no diagnostic value in Southern India, such, for example, as

the cephalic index and the facial angle. It is, however, worthy of note that the Todas, Kotas, and Badagas have the longest heads recorded by Thurston, and are in this respect separated from the Paniyans. The facial angle, though of great importance in separating prognathous from orthognathous races, is of little use as an aid to comparison and classification of the different communities of Southern India, in whom the average of the angle of Cuvier (with its vertex at the edge of the incisor teeth) ranges in the people examined by Thurston between 67° and 71°. Here, again, only the Badagas reach 71°, only one other group besides the Kotas reach 70°, and only the Sheik Muhammadans and the Paniyans fall as low as 67°.

Existing materials do not enable us to prosecute the analysis much further, but among the non-Aryan tribes of Southern India it appears as if we could trace two groups: (1) a taller, with moderately long arms, with long heads, and distinctly dolichocephalic, a moderate facial angle, and a mesorhine nose; (2) a very short, long-armed group, dolichocephalic, more prognathous, and with a very platyrhine nose. The Nilgiri Hill tribes are typical examples of the former group, and the dark-skinned, curly-haired Paniyans of the latter group.

It is a common belief among the European planting community that the Paniyans are of African origin, and descended from ancestors who were wrecked on the Malabar coast. This theory, which is based solely on their general appearance, breaks down on investigation. Of their origin nothing definite is known. An interesting account of these people is given by Thurston, and it is evident that we are dealing with a very primitive group of mankind, who have left traces of their former greater extension in the broad noses which occur among the lower Hindu castes.

In his earlier work Dr. Collignon, like other French an

thropologists, was inclined to place great value on the nasal index of the living as a distinguishing character of the races of Western Europe. Taking the three main constituents of the French nation, he found' the nasal indices came out in the following order:

Kymri....

Mediterraneans.

Celts'

63.39

65.48

67.20

And he naturally thought this was an excellent means of further distinguishing between the tall, dolichocephalic, fair race that came from the north, the fairly short, dolichocephalic, dark race of the south, and the short, brachycephalic, dark race of the centre.

In the same paper (p. 508) Collignon formulates the law that "in a given race leptorhiny is in direct relation to stature; the more it is raised, the longer the nose, the lower it is, the more the nose tends to mesorhiny."

As a result of his later researches Collignon finds that the nasal index in the living is of little practical value in French anthropology. He says ':

"Like the stature, the nasal index has received a serious blow (from being in the front rank for the classification of European races). It is true that this character maintains its incomparable value for the separation of the main trunks of mankind, but, so far as concerns the European races, properly so-called, it is incontestable that its value is diminished. The ethnic scale, which is naturally of a narrow range between races so allied to one another, may be neutralised by local variations in height. It thus

1 R. Collignon, "Étude anthropométrique élémentaire des principales Races de France," Bull. Soc. d'Anthrop. de Paris, 1883, p. 502.

2 This term is used in Broca's sense. See p. 121.

R. Collignon, "Anthropologie de la France; Dordogne," etc., Mém. Soc. d'Anth, de Paris (3), i., 1894, p. 43.

happens that the importance of the index is only relative, and that it loses the character of precision which we formerly credited it with."

We must now pass to a consideration of the nose from a craniological point of view. In the skull we find that the prominent part of the nose is formed of two elements, the nasal bones, and the upper jaw or maxillary bones; the former constitute the bridge of the nose, the latter bound the lateral and inferior margin of the nasal aperture, and they also flank the nasal bones so as to separate them from the orbits. The nasal bones are bounded above by the frontal or bone of the forehead. The mid-point of the fronto-nasal suture is termed by anthropologists the nasion; the corresponding spot in the living nose is the root. The nasal aperture is technically called the apertura pyriformis, the lower border of which has certain characteristics to which reference will be made. In the middle line of this lower border there is usually a bony projection, the nasal spine, which is continuous with the gristly and partly ossified nasal septum. There is no need to refer to the gristly portions of the external nose, as these are macerated away in dried skulls, and, though a description of them is here omitted, it must be remembered that they support and give the form to the nose as seen in the living subject.

The height of the nose is the line joining the nasion to the corresponding point at the base of the nasal spine. The breadth is the greatest diameter of the nasal aper

ture.

The cranial nasal index is the ratio of the nasal breadth to its height; this is obtained by multiplying the former by one hundred and dividing the product by the latter. The indices are grouped by Broca in a threefold classification into broad, medium, and narrow noses, the figures being:

The nasal indices to the left are those that were first suggested by Broca, and which have been subsequently adopted by French and English anthropologists; to the right are those in general use in Germany, as accepted at the "Frankfurter Verständigung."

In the following table I have collected from Broca, Topinard, Flower, and other sources, a selection of nasal indices. It is evident that we have here the same story that is told by the nasal index of the living. The black races are platyrhine, whether they come from Africa or Oceania. The yellow races, including the Indo-Polynesians and Americans, are mesorhine, and the European races are leptorhine.