6l2 



NA TURE 



[April 26, 1894 



sets of regions all differing in their names and in their bound- 

 aries? It would be like comparing the structures of different 

 animals as described in the works of a number of anatomists 

 each of whom had a different classification and a different set of 

 technical terms, so that before a single comparison could be 

 mude the terms used in one description would have to be trans- 

 lated into the ternis used in the other. In the study of geo- 

 graphical distribution, should this system prevail, the student 

 would find it necessary to adopt some one set of regions for his 

 own use, and then endeavour to translate the facts given by 

 each specialist into terms of that set before he could obtain any 

 clear conception or accurate knowledge of their comparative 

 distribution. 



An idea seems to be prevalent among biologists that there is 

 some lazv of distribution, that may differ for different groups, 

 and that may require different regions to exhibit it or to conform 

 to it. This, however, is a mere supposition ; but, if it is a correct 

 one, we shall certainly not be likely to disc.ver the "law " by 

 recording the facts of distribution in such a way as to render a 

 coinparati7j: i,\.vAy of ihem as difficult as possible. Laws of dis- 

 tribution can only be arrived at by comparative study of the 

 different groups of animals, and for this study we require a 

 common system of regions and a common nomenclature. 



It appears to me, however, that the " law," or at all events 

 the general principles on which the diversities of distribution 

 among land animals depend, is already fairly well understood. 

 What we require is to be able to work out the details in the 

 different groups, and thus explain certain difficulties or anomalies. 

 To detect anomalies it is essential to compare the distribution 

 of the different groups by means of a common system of regions. 

 If we construct regions to fit each group, the student of each 

 separate group will be apt to forget that it presents any anomalies 

 which require explanation 



Before leaving this part of the subject it may be well to give 

 a short account of the reasons which led to the original es- 

 tablishment of the six Sclaterian regions for the purpose of 

 facilitating the study of the geographical distribution of animals ; 

 in order to show that they are not arbitrary divisions, but 

 are founded on a large body of observations. It is evident, in the 

 first place, that many of the ordinary divisions of the geographer 

 serve well to define the areas characterised by special groups of 

 animals or plants. The South European, the Malayan, the 

 Brazilian, or the South African faunas and floras, are constantly 

 referred to, because those districts are really characterised by 

 distinct assemblages of animals and plants, and this un- 

 doubtedly depends partly on their possessing peculiarities of 

 climate resulting in peculiarities of vegetation — as forest, prairie, 

 desert, or woodland ; partly in their being limited by more or 

 less effective barriers, climatic or geographical ; and partly on 

 their past geological history and on the more recent changes of 

 physical geography they have undergone. But such areas as 

 these are too small and loo numerous to enable us to express 

 the broader features of the distribution of animals, and the 

 larger or primary geographical divisions — which were those 

 used by the older naturalists — are often unsuitable and mis- 

 leading, because they are not coincident in their boundaries 

 with those more permanent natural barriers which have mainly 

 determined the zoological specialities of different parts of the 

 globe. Yet some of the divisions of the geographer are such 

 well-defined and ancient areas that they do nearly coincide with 

 characteristic assemblages of animals ; and thus the geogra- 

 phical units, Europe, Asia, Africa, North America, South 

 America, and Australia, can be easily modified into six zoo- 

 logical regions, which do represent with considerable accuracy 

 the broad features of animal distribution. These modifications 

 may be briefly enumerated in order to show how the limits of 

 the regions have been arrived at. 



Beginning with Europe, we see at once that it is zoologically 

 homogeneous, since a large proportion of the species and all the 

 larger genera range over the whole of it. But the same genera, 

 in the case of the higher animals, at all events, prevail in North 

 Africa, mingled only with a few desert types, and we therefore, 

 for zoological purposes, add this area to Europe. It is interest- 

 ing to note that we have a clear explanation of this identity, 

 in the proofs that quite recently — that is, during the 

 Pleistocene period — Europe and North Africa were connected 

 both at Gibraltar, and from Sicily and Malta to 

 Tripoli, as indicated both by submarine banks which 

 still unite them, and by the fossil hippopotami and elephants of 

 the Maltese, Sicilian, and Gibraltan caverns. But further, if 



NO. 1278, VOL. 49] 



we go eastward from Europe into Siberia and Central Asia, we 

 find the same genera and many of the same species of mam- 

 mals and birds ranging all the way to the shores of the Pacific. 

 To such an extent is this the case that about fifty-six species of 

 British passerine birds range to Central and North-East Asia, 

 while no less than fifty-three species (or representative sub- 

 species) of land-birds are common to Great Britain and Japan. 

 Europe and North Asia are therefore parts of one zoological 

 region, the reason being that there is not, nor has been in 

 recent geological times, anv effective barrier between them, 

 while in climate they are sufficiently alike. 



Here, then, we have roughly marked out our first great zoo- 

 logical region — the Palsearctic, or northern old-world region. 

 Southward its limits are undefined, and where there are no 

 well-marked barriers, such as the Himalayas or the desert, 

 there will always be a greater or Jess width of border-land 

 between two conterminous regions. 



Now this Palsearctic region is, fortunately, the only oae that 

 differs very largely from the ordinary geographical quarters or 

 divisions of the globe. For when we go to Africa we find 

 that, leaving out the northern portion, which we have seen to 

 be essentially European, the remainder constitutes a very dis- 

 tinct and compact area zoologically — the land of giraffes, 

 zebras, hippopotami, baboons, and antelopes — which has been 

 termed the Ethiopian region. Then we have southern or 

 tropical Asia, together with the larger Malay Islands, which 

 we know must recently have formed a part of it, constituting 

 the Indian or Oriental region, and corresponding almost ex- 

 actly with tropical Asia. Then we come to Australia, which 

 forms the nucleus of another well-marked region, including with 

 it, however, most of the Pacific Islands, with New Guinea and the 

 Moluccas. Turning now to the western hemisphere, we have 

 South America and North America, which, with slight modifi- 

 cations, form two well-defined regions — the Neotropical, includ- 

 ing all South America with the tropical portion of North 

 America and the West Indian Islands ; the Nearctic, com- 

 prising the remainder of North America. 



Now, I do not think thit any one has denied that these are 

 truly natural divisions of the eaith from a broad zoological point 

 of view. The controversy respecting them has turned wholly 

 on whether they are of equal rank. This point, however, will be 

 referred to later on. We are now dealing with the question of 

 the need of other modes of dividing the earth's surface in order 

 to exhibit and to study the distribution of certain groups of 

 animals. I have already urged that to do so would defeat the 

 very object aimed at, and render the study of geographical dis- 

 tribution very much more difficult. I have shown how readily 

 the Sclaterian regions enable us to describe or tabulate the dis- 

 tribution even ot a group which has been said to "pay no 

 attention to them whatever" ; and I have now just pointed out 

 that these six regions are, admittedly, natural, which can only 

 be because during the more recent geological periods they have 

 formed single more or less continuous areas, while separated 

 either by geographical, climatal, or biological barriers from the 

 adjacent areas. 



Now the only real interest of the study of geographical dis- 

 tribution lies in its giving us a clue to the causes which have 

 brought about the very divergent and often conflicting distribu- 

 tion of the various species, genera and higher groups, and by 

 thus being able to explain most of the anomalies of distribution. 

 These causes we can trace, in many cases, either to geogra- 

 phical or climatal changes in the past, which temporarily 

 removed the barriers that now exist or interposed others that 

 are now absent : or, on the other hand, to the recent extinction 

 of groups in certain regions where they formerly abounded ; or, 

 again, to the very different powers of dispersal possessed by 

 different organisms, which enable some groups to spread easily 

 where others are stopped by an insurmountable barrier. Now 

 it is usually this last phenomenon, of varying powers of dispersal, 

 that has led the students of certain groups to urge that the old- 

 established regions do not serve their purpose. But when a 

 group can more or less easily traverse the barrier between two 

 regions, however permanent that barrier may be, the fact 

 enables us to explain the exceptional distribution of that 

 group, but it does not render the established regions less 

 natural, or require a fresh set of regions, which would certainly 

 not be natural in any broad sense, to explain them. 



Again, it will usually, perhaps always, be found that even itt 

 the groups appealed to as requiring a new set of regions, a por- 

 tion of the species, and even of the genera, are limited to the 



