;64 



NATURE 



[July 5, 1917 



OUR BOOKSHELF. 



Horses. By Roger Pocock. With an Introduc- 

 tion by Prof. J. Cossor Ewart. Pp. x + 252. 

 (London : John Murray, 1917.) Price 55. net. 

 This is an entertaining little volume, written by 

 one who has spent much of his life among horses 

 as a ranchman; and from the ranchman's point 

 of view he surveys .horsemanship the world over. 

 As a standard of comparison this is useful 

 enough, but, unfortunately, it has distorted his 

 judgment. Hence he is led to assure us that 

 "the pleasure horse and his equipment are so 

 highly specialised for running and jumping that 

 they have ceased to possess the slightest value 

 for civil and military working horsemanship." 

 Yet, as a matter of fact, a large proportion of 

 British cavalry has been horsed during the 

 present war by animals taken from the hunting- 

 field. The best bred of these animals, indeed, 

 are generally considered to make the finest 

 cavalry chargers in the world. The author has 

 some pertinent criticisms on our saddles and mode 

 of riding, and on our treatment of horses in 

 and out of the stable, which will at least repay 

 careful consideration from those immediately 

 concerned. 



Of his own feats in the saddle he has much to 

 say, and some of these indeed savour of the won- 

 derful. They are, at any rate, eminently read- 

 able. Less entertaining are his sneers at the 

 "scientist," whom he regards as "an 

 amateurish, unpractical sort of person, who 

 cannot either ride or cook " — these being the 

 only accomplishments for which he has any 

 regard. His confidence in his powers of obser- 

 vation, and his knowledge of the lie of the land, 

 even in unexplored country, are so absolute that 

 he has no use for either the compass or maps ! 



In a chapter on the origin of the horse — which 

 he owes to the " scientist "—he assures us that 

 "the bald skin of the pig is boldly painted in 

 splashes of pink and brown to imitate the lights 

 and shadows of forest undergrowth. The forest 

 ancestors of the horse were bald and painted in 

 the same way. . . ." Pink pigs may be seen 

 in our farmyards in plenty, but we know of no 

 wild race similarly coloured, and there is no 

 reason for supposing that the forest ancestors of 

 the horse were "bald." 



If the author had adopted a less superior atti- 

 tude his book would have been even more 

 readable than it is. 



Bacon's New Series of Physical Wall Atlases : 

 British Isles. Scale 1:1,187,000 (i8'7 miles 

 to an inch). (London: G. W. Bacon and Co., 

 Ltd.) Price 265. 



The seven maps in this series vary a great deal 

 in value. The orographical map, with layers in 

 two colours and showing also trunk railways and 

 Roman roads, the geological map, and the rain- 

 fall map are all clear and useful. The isotherm 

 map would be improved by the omission of the 

 mean annual isotherm, which is not only confus- 



NO. 2488, VOL. 99] 



ing when on the same map as the January and 

 July ones, but of little or no value in geo- 

 graphical teaching. The map showing vegeta- 

 tion and productions is not -a success. The large 

 letters to indicate the location of various 

 industries are crowded and somewhat arbitrarily 

 selected. Thus Aberdeen is given no granite 

 industry, while Ballater is ; Leeds has no indica- 

 tion of its leather factories, or Elgin of its dis- 

 tilling and brewing. The West Riding coalfield 

 extends much further east than shown. In the 

 population map the many colours employed give 

 a bad impression and make a confused map. 

 The last map, that of communications, might 

 well have been omitted. It shows some of the 

 lines of the various railway companies all dif- 

 ferentiated from one another. There seems to 

 be little object in teaching the ownership of each 

 line, but the great objection to the map is the 

 want of any indication of relief. Without this 

 there is no sense in teaching lines of communi- 

 cation. Moreover, the orographical map does 

 all that is required in this matter. 



All the maps have the same names printed in 

 ground colour for the use of the teachers. It is 

 useful to have the series all on the same scale, 

 but we are at a loss to understand why this par- 

 ticular scale should have been selected. It does 

 not facilitate comparison with maps on other 

 scales. 



Meteorologie du Bresil. By C. M. Delgado de 

 Carvalho. Pp. xix + 527. (London : John 

 Bale, Sons, and Danielsson, Ltd., 1917.) Price 

 255. net. 



The publication of this work is very welcome,, 

 not only to meteorologists, but also to those 

 having interests in this progressive Republic. Few • 

 except those who have had occasion to deal with 

 South American meteorological observations can 

 appreciate the onerous nature of the task which 

 the author set himself in the preparation of this 

 comprehensive climatology of his native country. 

 In most cases the difficulties have been successfully 

 surmounted, with the result that we have put be- 

 fore us in a very readable and instructive form a 

 series of pictures showing not only the diverse char- 

 acter of the climates of Brazil, but also the inter- 

 connection that exists between climatic conditions 

 and migration, immigration, and public health. 

 The work opens with a summary of the broad 

 climatic features, and of the local and general 1 

 conditions affecting them. An interesting section | 

 deals with the " action centres " of the atmosphere, 

 and of the genesis of the tropical rains. Nearly 

 two-thirds of the book are devoted to an analysis 

 of climatological data, this section including no 

 fewer than thirty-four separate studies of local 

 climate, the stations selected ranging from Para, 

 near the equator, to Pelotus, in lat. 32° S. In some 

 areas, such as the State of San Paulo, where 

 stations are numerous, much additional information i 

 is given, especially regarding the diurnal range of j 

 the climatic elements. The unique position which ! 

 Brazil occupies for a study of various meteoro- < 



