BY CHARLES HEDLEY. 393 



all, it is assumed that if the depth is over a certain amount — say 

 1,000 fathoms^former land connection was not possible; then 

 ■comes the study of the flora and fauna of those islands which are 

 thus situated, and those are then looked upon as characteristic of 

 such islands — other islands have these characteristics — the con- 

 clusion is drawn that they also have never been connected with 

 the land."* 



The only safe mode of reasoning is to eliminate the factor of 

 the depth of intervening seas, since we cannot count the amount 

 of past possible upheaval or depression, and to rely on biological 

 data alone. Lest the important distinction between what Baur 

 so happily terms "harmonic" and "disharmonic" faunas, should 

 escape attention, I venture to again express it thus. 



Let an area of say ten square miles be selected in, for example, 

 England ; a census of its fauna will yield a certain total of 

 mammals, frogs, birds, fish and so on. Let another such census 

 be made of an equal area in the United States and in Australia. 

 Now though the individual species would be different in each of 

 the three resulting catalogues, yet a general harmony in the pro- 

 portion of each group to the whole will prevail. On contrasting 

 the totality of the Central Pacific fauna with such lists, the gaps 

 at once make discord. For example, the loss of the mammals, 

 snakes and amphibia, leaves the whole Phylum Chordata with 

 scarcely a representative in the Central Pacific. 



The deduction from this comparison is that the population of 

 the Central Pacific has been received by drift, from a continent 

 or continental islands. Consequently all that the atolls have, 

 their source should have also. But of the population of that 

 source, only such may extend to the atolls as may first endure the 

 ordeal of transit, and secondly obtain the means of life upon 

 arrival. And the disharmony will result in the elimination from 

 that atoll fauna, as compared to the continental, of the animals 

 which have failed to comply with these two conditions. 



* Deane— Proc. Lmn. Soc. N.8.W. xxi. 1896, p. 847. 

 26 



