458 



NATURE 



[September 17, 1896 



Again, the " Austro-Malayan region " of Mr. Lydekker 

 is merely a border-land between the Australian and 

 Oriental regions, and has no sort of claim to the rank 

 here assigned to it. It has few, if any, indigenous types of 

 mammals, and cannot for a moment be put on a par, 

 as it is in Mr. Lydekker's scheme, with the Ethiopian 

 region, which has numerous families, both of mammals 

 and birds, restricted to its area. Mr. Lydekker's "Austro- 

 Malayan region," except Celebes — which is certainly a 

 difficult subject — may be safely anne.xed to the Australian 

 region. Celebes has been also referred there by Mr. 

 Wallace, but on the whole we opine that it would be 

 better placed as a distinct sub-region of the Oriental 

 region. 



We now come to Mr. Lydekker's "Arctogasic realm," 

 or " Arctogasa," as we prefer to call it. This is divided 

 by Mr. Lydekker into five regions, as shown above. As 

 regards the separation of Madagascar and its islands 

 from the Ethiopian region, under the name of the 

 " Malagasy region," there is much to - be said in its 

 favour, and we do not deny that our author has some 

 good grounds to go upon. It is obvious that the mammal 

 fauna of Madagascar, as well shown by Mr. Lydekker 

 (see p. 215 of his work), is one of the most extraordinary 

 on the world's surface— not only for what it has, but still 

 more, perhaps, for what it has not. We can, therefore, 

 offer no serious objection to Dr. Blanford's proposal 

 (accepted by Mr. Lydekker) to raise the rank of 

 Madagascar from that of a sub-region (as it has been 

 treated by Messrs. Sclater and Wallace) to that of a full- 

 blown region. 



Mr. Lydekker's Ethiopian and Oriental regions 

 remain much the same as Mr. Wallace's ; but as regards 

 the next two— the Holarctic and the Sonoran, there is a 

 wide difference. Mr. Lydekker, misled by Dr. Merriam 

 and other American writers, who take a narrow view of 

 the subject, proposes to annex the northern part of 

 America to the northern part of the Old World, calling 

 it altogether Holarctic ; while the more southern part of 

 North America, down to the boundaries of the Neo- 

 tropical region, is denominated "Sonoran." To assent 

 to this proceeding, however, would only involve us in 

 further difficulties. Most of the " Sonoran " mammals 

 penetrate far into the north, outside its supposed limits. 

 On examining the list of the Sonoran types (p. 379) and 

 that of the "Western Division of the Holarctic region" 

 (p. 344), we shall find them meagre indeed, and quite 

 insufficient to support a distinction between two regions. 

 The polar area may, in fact, be safely regarded as border- 

 land between the PaL-earctic and Nearctic regions. It 

 must be recollected that Northern America was, in com- 

 paratively recent days, covered by the polar ice-sheet even 

 much more extensive than that of Northern Europe. This 

 destroyed nearly all animal life, and drove most of the 

 remainder into Mr. Lydekker's "Sonoran region." On 

 the disappearance of the ice-sheet the northern land was 

 naturally repeopled from the adjacent part of Asia 

 across Behring's Straits, as well as from the Sonoran 

 region. Hence, no doubt, came such characteristic 

 Pal^arctic forms as the sheep, the bison, the mountain- 

 goat, and the stag into North America. Of these, how- 

 ever, all but the mountain-goat have penetrated into the 

 Sonoran region, and we have some doubts whether 

 NO. 1403, VOL. 54] 



Haploccros is not likewise to be met with within its sup- 

 posed boundaries. 



Another serious objection to the " Holarctic region " is 

 that, as regards birds at least, by adopting it we shall mix 

 up some of the most characteristic forms of the New World 

 in the same primary division as those of the Old World. 

 Take, for example, the humming-birds — a most eminently 

 typical group of the New World. Humming-birds range 

 all over Canada in the summer, and on the west of the 

 continent pass up to Alaska. Following Mr. Lydekker^s 

 scheme, we should have to place the Trochilida; in the 

 " Holarctic list." The same would be the case with the 

 Mniotiltine warblers {Mniotiltida)^ the greenlets {Virc- 

 onida), the hang-nests {Icteridcc), and other forms which 

 are absolutely restricted to America, and utterly foreign 

 to the Old World [PaUcogean) avifauna. 



On the whole, therefore, we cannot doubt that Mr. 

 Lydekker would have been more prudent to stick to the 

 old-fashioned " six regions." Even had he not quite 

 agreed to them, he might have sheltered himself under 

 Mr. Wallace's authority,and safely followed his leadership. 



In his intimate acquaintance with fossil mammals, Mr. 

 Lydekker had, as we have already stated, a great 

 advantage over his fellow-workers in the same field, and 

 one of which he has not failed to make good use in some 

 of his arguments. This branch of the subject is certainly 

 much more completely discussed in the " Geographical 

 History of Mammals" than in any other work of the 

 same character, and we are duly grateful to the author 

 for the many novel facts he has thus set before us. At 

 the same time it should be recollected that, w^hile we 

 are pretty well acquainted with the present mammal- 

 fauna of the earth and the facts of its distribution, we 

 know comparatively little about the past. The " im- 

 perfection of the geological record " should be always in 

 our minds when arguments are used taken from the little 

 that is yet known of the ranges of extinct mammals, our 

 notions of which may in many cases come to be seriously 

 modified by discoveries yet to be made. 



On the whole, however, we must allow that Mr. 

 Lydekker's volume forms a valuable contribution to the 

 " Cambridge Geographical Series," and that the general 

 editor has done wisely in securing such a well-written 

 essay on this branch of his subject from a palieontological 

 point of view. Although we notice a few typographical 

 errors, the volume is well printed, and excellently illus- 

 trated by numerous process-blocks introduced into the 

 text, and by a chart of the zoological regions. Altogether 

 it contains a large mass of information reduced into a 

 small compass, and will meet, we are sure, with generous 

 appreciation from all students of distribution. 



THE RATIONAL STUDY OF PLANT- 

 DISTRIBUTION. 

 Lehrbuch der Okologischen Pflanzengeographic eine 

 cinfiUu'ung in die kenntniss der P/lanzenvereitie. Von 

 Dr. Eugen Warming. Deutsche Ausgabe von Dr. E. 

 Knoblauch. (Beriin : Gebruder Borntraeger, 1S96.) 



AN account of the principles underiying the facts of 

 the geographical distribution of plants has long 

 been a desideratum. Although various persons ha\e 

 written on the subject, they have not, for the most pan. 



