740 



SCIENCE. 



[N. S. Vol. III. No. 72. 



of animals, since eacli group is subject to differ- 

 ent laws. Thus a division obtained by the 

 study of the prevailing conditions in one group 

 is often exactly the opposite of that found to 

 prevail in other groups. From this disagree- 

 ment arose the continu.ous dispute between dif- 

 ferent writers with regard to the number and 

 the limits of the zoogeographical divisions, each 

 wishing to transfer the results obtained in his 

 favorite group to other groups. 



3. The actual distribution of animals is the 

 result of development during the course of the 

 geological history of the earth. While many 

 animals show a distribution which corresponds 

 to the physical conditions of recent times, many 

 others point clearly to conditions of former 

 periods, and their distribution is only intelligi- 

 ble under the supposition that formerly differ- 

 ent conditions prevailed on the earth. 



Thus we should expect that investigations 

 founded on the actual distribution of animals 

 are in the first place incomplete, and in the 

 second the results obtained are contradictory 

 in many cases. In order to overcome the latter 

 difficulty, statistical lists of the distribution of 

 these animals have been prepared showing 

 which distributional features are most common. 

 But I object even to these statistics. My first 

 reason shows clearly that such statistics never 

 are complete, and it is very dangerous in science 

 to rely upon statistics deficient in the main 

 quality by means of which they are useful at 

 all. 



From these considerations I am induced to 

 use the deductive method, and to construct zoo- 

 geographical divisions according to the differ- 

 ences in the physical conditions influencing the 

 distribution of animals. But I remark ex- 

 pressly, I do not regard such a division of the 

 earth as the final aim that should be reached in 

 zoogeography, but only as a means which facili- 

 tates zoogeographical study. My divisions rep- 

 resent only a rough sketch of the distribution of the 

 different conditions of life in recent time. Of 

 course, these divisions do not agree with those 

 assigned to animals the range of which is due 

 to conditions belonging to former times; but 

 even in such cases my divisions have a decided 

 advantage. If there are any exceptions in the ac- 

 tual distribution of certain forms we see at once 



that these animals do not follow the general 

 rules according to which the divisions are con- 

 ceived, and the knowledge that certain laws do 

 not control particular cases is a considerable ad- 

 vantage in revealing the true causes of these 

 peculiarities. For the whole point or aim of 

 zoogeographical research is to find out the causes 

 of the distribution of each animal form. 



The above reasons, I think, are sufficient to 

 demonstrate that my starting point has certain 

 advantages over that of other students in zo- 

 ogeography. Notwithstanding the results of my 

 investigations are very similar to these obtained 

 by Prof Gill. This is due, I believe, to the ex- 

 tensive and correct character of his preliminary 

 work, to the exact and fundamental study of 

 the actual distribution of certain groups of ani- 

 mals, and to the full use he has made of the 

 known facts. On the other hand, I think. Prof. 

 Gill's method is not so fandamentally different 

 from mine as it seems to be perhaps according 

 to his own statement. It is true he ' prefers 

 the inductive method ' (p. 515), and his divis- 

 ions are adapted in some degree to the actual 

 distribution of certain animals; nevertheless his 

 chief marine divisions are conceived according 

 to a physical principle, to the temperature of the 

 ocean waters, a principle which was first intro- 

 duced by Dana, and the importance of which is 

 recognized by Prof Gill in the concise sentence: 

 "Temperature is a prime factor, and land a sec- 

 ondary, in the distribution of marine animals."* 

 On this point our opinions agree completely, 

 and thus, I think, our starting points are not so 

 extremely different, since Prof Gill in construct- 

 ing his zoogeographical divisions of the seas pays 

 due attention to temperature, which is at least 

 one, and indeed the most important, physical 

 factor. 



With regard to the objections of Prof. Gill to 

 my life districts, I should like to add here that 

 I do not fully understand why he says they are 

 misconceived, since they are framed in contra- 

 vention of my own principle of continuity. If 

 all the life districts were continuous, any further 

 divisions would be impossible and needless, as 

 is the case in the abyssal (bassalian) district, 

 and even the discontinuity of the others obliges 



*Presidential Address Biol. Soo. 'SVasbiDgton, Jan. 

 19, 1883, p. 39. 



