April 8, 1892.] 



SCIENCE. 



203 



But the most curious statistics are those relating to the 

 ■nasal indices of the tribes examined. They corroborate the 

 high value of this physical element in racial anatomy. The 

 nasal index is found in India in two widely distinct types; 

 the one platyrhine to a degree closely approaching the negro 

 (88-95), the other leptorhine about in the same proportion as 

 in western Europe (67-72). These indices bear a constant 

 relation to the order of social precedence, to the distinctions 

 of caste, and to the organization of the family. " It may 

 be laid down as a working hypothesis, if not as an absolute 

 law, that the social position of a caste varies inversely as its 

 nasal index." Everywhere the narrow-nosed Brahmin is 

 at the top, the broad-nosed Pariah at the bottom. Wherever 

 there is a high index, — above 80, — we find a low social po- 

 sition and the totemic sub-division of the tribe; wherever the 

 index is low, — below 75, — we are equally sure to meet high 

 rank and an eponymous family system. 



Incidentally it may be added that these investigations bear 

 out the ancient Indian traditions that the Aryan nations of 

 India entered the peninsula from the north-west, and de- 

 stroyed or subjugated the ancestors of the dark, flat-nosed 

 Kols. the "snub-nosed blacks," often referred to in the 

 ancient Vedic war-songs. 



OSTEOLOGICAL NOTES. 



In previous papers (Science, Vol. xvi., p. 332, Vol. xvii., p. 

 117, Vol. xviii., p. 53) we have assumed that the modifications 

 presented by the jugal arch in the Mammalia are due to the 

 various influences derived from use or disuse, correlated 

 necessarily with the habits and environment of the animal. 

 In no order is the specalization of the arch, under the influ- 

 ences of natural selection, more clearly exhibited than in the 

 Insectivora. 



Adopting the classification of the highest authorities, and 

 notably that of Dr. Dobson, this order may be divided into 

 two sub-orders, first, Dermoptera, embracing only one species 

 — Galeopithecus volaus — and, second, Insectivora Vera, 

 ■which comprehends all the remaining families. This second 

 sub-order may be divided in turn into two groups. In the 

 first, — including the families. Tupaidoe, Macroscelidoe, 

 Erinaeeidoe, Talpidoe, and Soricidoe, — the true molars have 

 W-shaped crowns. In the second group, including the 

 Centetidce, Solenodontidoe, Potamogalidoe, and Chrysochlo- 

 ridoe, these same teeth have V-shaped crowns. 



Accepting the above classification, the Insectivora, so far 

 as concerns the jugal arch, may be brought into three 

 groups. 



1. Those in which the arch is complete and well developed, 

 comprising the Tupaidoe, Macroscelidoe, Ehynchocyonidoe, 

 Galeopithecidce. 



2. Those in which the arch is complete but more or less 

 feebly developed, comprising the Erinaeeidoe, Talpidoe, 

 Chrysochloridce. 



3. Those in which the arch is partially or wholly deficient, 

 comprising the Centetidce, Potamogalidoe, Solenodontidoe, 

 Soricidoe. 



The Tupaia (Squirrel-shrew) may be taken as a typical 

 form of the first group. The jugal arch is well developed, 

 a post-orbital process from the frontal meeting a correspond- 

 ing one from the malar, thus forming a complete bony or- 

 bital ring. The malar has a large longitudinal oval vacuity, 

 which, although unique in this case, when taken with simi- 

 lar vacuities in the palate of this genus, as also in some of 

 Ihe other Insectivora, points unmistakably to the Marsupilia. 



The horizontal curvature of the arch is sufficient to counter- 

 act any inherent weakness due to the vertical curvature 

 with its convexity downwards. The temporal fossa is 

 moderately extended, while the coronoid surface of the 

 mandible presents a large backward projecting surface rising 

 high above the transversely produced condyle. 



In the second group, where the arch although complete is 

 for the most part weak, the cranium presents marked modi- 

 fications. In Erinaceus and Gymnura the arch is formed 

 mostly by the processes of the Squamosal and maxilla which 

 join, while the molar is very small and occupies in a splint- 

 like form the outer and under sides of the centre of the arch. 

 There- are no traces of any post-orbital processes. The tem- 

 poral fossa is deep and extended, while aMitional surface is 

 afl'orded for the temporal muscle by the prominence of the 

 sagittal and occipital crests. The ascending ramus of the 

 mandible with its broad concave coronoid surface and the 

 development of the pterygoid fossae denote increased mastica- 

 tory powers, in spite of the apparent weakness of the but- 

 tress. 



In the Talpidoe, certainly in all of the truly fossorial of 

 the family, the jugal arch is slender and exhibits no distinct 

 malar bone, no occipital or sagittal crests, and no post-or- 

 bital processes. The mandible is long and the vertical por- 

 tion presenting a moderately extended coronoid surface with 

 a small transverse condyle. The infra-orbital foramen is of 

 great size, being a very slender osseous arch which serves 

 for the transmission of the large infra-orbital branch of the 

 trifacial, which affords the necessary supply of sensory 

 nerves to the muzzle. 



In the Chrysochloridce (Golden moles), which in the gen- 

 eral shape of the skull present modifications different from 

 all other Insectivora, the jugal arch is in some species so ex- 

 panded vertically, that, as Dr. Dobson remarks, " their up- 

 per margins rise above the level of the cranium giving addi- 

 tional origin to the large temporal muscles." There is no 

 post-orbital process given off either by the frontal or zygo- 

 matic arch. As regards the mandible, the coronoid process 

 is little elevated and in some species is nearly level with the 

 transversely extended condyle. 



In the third group the arch is incomplete, and in one in- 

 stance, at least, may be described as entirely absent. lu the 

 Centetidce, the skull is long and narrow, and marked by 

 largely developed occipital and sagittal crests which serve as 

 attachments for the muscles of temporal origin. The zygo- 

 matic processes of the maxilla and squamosal are very short 

 and rudimentary, while the malar is entirely absent. The 

 temporal fossae are very large, and the skull retains nearly 

 the same width at their anterior and posterior regions. There 

 is not a trace of a post-orbital process. The infraorbital 

 foramen is circular, and capacious. There are no pterygoid 

 fossae. The coronoid process of the mandible is largely de- 

 deloped, its inner surface being concave, and its outer sur- 

 face flattened. The condyle is small and circular, while the 

 glenoid surface is transversely concave. 



The other families of this group with the exception of the 

 Soricidoe agree with the Centetidce in the modifications of 

 the skull that have been described. In the Soricidoe the 

 cranium is broadest just behind the glenoid surfaces. There 

 is no jugal arch and no trace of a post-orbital process. Fre- 

 quently there is present a strongly marked lambdoidal ridge 

 as well as a sagittal crest. There is no pterygoid fossa, but 

 very large vacuities exist on each side of the basis cranii. 



The mandible resembles that of the Talpidoe, although the 

 horizontal ramus is shorter, while the ascending one "pre- 



