1896. SOME NEW BOOKS. 393 



cussed at considerable length, and Mr. Lydekker comes to the 

 conclusion that the Didelphidae and Dasyuridae probably originated 

 from a common stock in South-Eastern Asia about the end of the 

 Cretaceous period. The members of the first-named family after- 

 wards migrated into Europe and Asia, thus accounting for their 

 sudden appearance in the Oligocene deposits of those areas, while the 

 Dasyuridae, on the other hand, passed into the Australian region, 

 where, being isolated, they gave rise to the highly differentiated and 

 specialised Marsupial fauna, including the Diprotodontia. The 

 differentiation of the latter probably took place at a very early date, 

 since among the Australian types of Marsupials lately discovered in 

 the Miocene of Patagonia, certain forms already exhibit the dipro- 

 todont modifications in the lower jaw. It is, of course, possible, and 

 perhaps even probable, that this modification may have arisen inde- 

 pendently in the S. American forms, but in any case it is hardly 

 likely to have occurred there sooner than in the main centre of the 

 group, Australasia. The former existence of an Antarctic continent, 

 the relations of the Ethiopian fauna to that of S. America and of 

 Madagascar, and many other questions of like interest will here be 

 found thoroughly threshed out, all available evidence being brought 

 to bear upon them. 



In short, it may be said that this work is a very valuable 

 contribution to the study of distribution, and should be read by all 

 who are in any way interested in such problems. At the same 

 time, it is not without defects, the more to be regretted, because in 

 most cases they could easily have been avoided. For instance, in 

 Fig. 12, the fact that the feet there represented are certainly not those 

 of the elephant and hyrax, as stated in the legend, will be at once 

 obvious to everyone, particularly to the author. Again, the statements 

 that " the Soricidae or Shrews are represented in Ethiopia only by 

 three species," and that the Potamogalidae are represented in 

 Madagascar by Microgale (p. 234), are incorrect. In the last case 

 Geogale is clearly intended, and, indeed, on p. 219 it is described as 

 being a Malagasy representative of Potamogale. 



The printing and illustrations are, on the whole, good, but the 

 mingling of wood-cuts and process-blocks has a somewhat unpleasing 

 effect. 



The Geography of Plants. 



Manuel de Geographie Botanique, par Oscar Drude, traduit par Georges 

 Poirault, at revu et augmente par I'auteur. Livraisons 11-13, pp. 401-512. 

 Paris: Klincksieck. Prix de chaque livraison, i fr. 25 ct. 



We are glad to see that the French translation and edition of Oscar 

 Drude's useful work on Plant Geography is so nearly finished, and 

 shall hope soon to have the opportunity of reviewing the work as a 

 whole. Meanwhile, we may mention that the present issue brings us 

 nearly to the end of Part V. " The regions of vegetation of the earth 

 arranged in geographical order." It contains the concluding pages 

 of Chapter ii. on " Northern Floral Regions," the whole of Chapter iii. 

 on " Tropical and Southern Floral Regions," while Chapter iv., on 

 the " Oceanic Floral Region," begins on p. 511. The bibliography 

 at the beginning of each section is a useful feature, and we would 

 suggest that every effort be made to render it as complete as possible. 

 A good deal of work has recently been done both in this country and 

 in Germany on the flora of Tropical Africa ; but while reference is 

 made to that of Prof. Engler and his colleagues, work done at Kew 

 and the British Museum during the last few years is almost ignored. 



2 E 



