3IO 



NATURE 



[November i8, 1915 



print; but Mr. Crooke has introduced many im- 

 provements. He has brought the statistics up to 

 date, secured greater uniformity in spelling the 

 vernacular names, added numerous footnotes, cor- 

 recting statements and theories which have been 

 proved to be erroneous or doubtful, and given an 

 appreciative memoir of the author's pioneer 

 anthropological work in India. 



Important additions also have been made to the 

 illustrations, which previously had been restricted 

 to rather inferior reproductions of several of the 

 beautiful photographs by Sir B. Simpson in 

 Dalton's "Descriptive Ethnology of Bengal" — a 

 work also now difficult to procure. The inclusion 

 of some excellent photographs for other parts of 

 India makes the series now more representative. 



The main thesis of the book is the contentious 

 one on the origin of the caste-system. There is no 

 doubt that Risley exaggerated the antiquity of the 

 present social grouping of the people in his belief 

 that the supposed fixity of caste dated to a 

 remotely early period, and had thus presumably 

 preserved in India a remarkable purity in physical 

 type. Mr. Crooke, however, shows that caste in 

 its modern rigid form in India is relatively recent. 

 The older custom, as seen in the Vedas and Epics, 

 recognised the possibility of a member of the 

 "warrior" class (Kshatriya) becoming a Brah- 

 man, or vice versd. A second wife could be taken 

 from any lower class, and such laxities in practice 

 still prevail in the more outlying districts of the 

 Himalayas and the Panjab amongst groups of 

 relatively pure Aryan stock. In many areas in 

 India it is shown that the existing tribes and 

 castes represent mixtures of various races which 

 had amalgamated within a comparatively late his- 

 toric period. 



On the disputed origin of the warlike castes of 

 Upper India, the long-headed Rajputs, Jats, and 

 others, Risley opposed the present fashionable 

 theory which would bring these people from Cen- 

 tral Asia. This he did on the old idea that the 

 people of Central Asia were of a uniform brachy- 

 cephalic type, which is now known to be certainly 

 not the case. Mr. Crooke seems to approve the 

 recent theory of Smith and others that the Gurjari 

 and the other associated " Rajput " tribes of 

 Upper India are largely, if not wholly, formed by 

 the Hun invaders of the early centuries of the 

 Christian era. Yet it is precisely in these and 

 associated tribes that Risley finds his purest 

 Aryan type ! 



Regarding the Marathas, the dominant tribe of 



Western India, Risley 's suggestion, based upon 



some reputed, but not clearly established, brachy- 



cephalism in the Deccan, is that they originated 



NO, 2403, VOL. 96] 



in bodies of Scythians driven down from the 

 Western Panjab and intermarrying with the 

 Dravidians. Mr. Crooke points out that there 

 is no historical nor even traditional evidence of 

 any Scythian migration into the Deccan ; whereas 

 the Marathas are closely connected with a mixed 

 race of cultivators extending over a wide area, 

 from the Deccan to the Ganges Valley, and known 

 as Kumbi or Kurmi. These and other results of 

 later research would no doubt have been con- 

 sidered by the author himself had he lived. 



To many workers the most important part of the 

 book will perhaps be the anthropometric tables, 

 consisting of seriations of the several physical 

 types. These and the maps of caste-distribution 

 are of permanent value. The large collection of 

 caste-proverbs will be found curious and interest- 

 ing to students of folk-lore. 



The only mistake we have noticed in the refer- 

 ences is that the article cited at page 2 from the 

 "Journal, Royal Asiatic Society" should be for 

 the year 1898, and not 1908. 



L. A. Waddell. 



PHILOSOPHY OF SCIENCE. 



(i) Proceedings of the Aristotelian Society. New 

 series, vol. xv., containing the papers read 

 before the Society during the Thirty-sixth Ses- 

 sion, 1914-1915. Pp. 441. (London : Williams 

 and Norgate, 1915.) Price 105. 6d. net. 

 (2) Selections from the Scottish Philosophy of 

 Common Sense. Edited by G. A. Johnston. Pp. 

 vii + 267. (Chicago and London: The Open 

 Court Co., 1915.) Price 35. 6d. net. 

 (i) /^~\F the twelve papers and symposia col- 

 V_y lected in the Proceedings of the Aris- 

 totelian Society for the last session, four are direct 

 criticisms of positions taken up by Mr. Bertrand 

 Russell. In the inaugural address on " Science 

 and Philosophy," Dr. Bosanquet criticises the 

 view, maintained in Mr. Russell's recent Lowell 

 Lectures, that philosophy, as the science which 

 aims at stating all that can be known a priori 

 about all possible worlds, should be ethically 

 neutral, and that it is just because philosophy in 

 the past has been biased by the desire for agree- 

 able conclusions that philosophy has not made the 

 same progress as the physical sciences. Dr. 

 Bosanquet holds that this view implies an ante- 

 cedent limitation of philosophy, and involves the 

 confusion that because the interest of philosophy 

 is purely theoretical, therefore its subject-matter 

 is itself theory and its objects. 



In a paper of extraordinary acuteness, Mr. 

 C. D. Broad deals with Mr. Russell's attempt in 



