8 CLARK: RECENT CRINOID FAUNA 



of these animals from littoral waters having a seasonal fluctuation 

 of temperature and salinity, and because of the complete inter- 

 graduation of the littoral and the abyssal types. 



Yet the very features which place the greatest obstacles in the 

 way of outlining the zoogeographic divisions indicated by the 

 crinoids at the same time suggest that these divisions are of more 

 than usual significance, and are more fundamental in character, 

 especially in their relation to the zoogeographic divisions of the 

 past, than those of the other groups of marine organisms. 



In the construction of the scheme outlined below, instead of 

 following the usual method of subordinating the biological to 

 the geographical aspect of the problem, I have examined all the 

 known species of recent crinoids from the point of view of their 

 systematic affinities and their obvious relationships, later assign- 

 ing them to the various zoogeographic divisions in which they 

 appear naturally to group themselves, so that these divisions are 

 outlined purely from the biological viewpoint, and follow, if 

 it may be so expressed, the phylogenetic rather than the geo- 

 graphical migrations of the class. 



The crinoid fauna of the present seas is found to be divisible 

 into two main sections, which, though faunally equivalent, are 

 different in size and range. These two zoogeographic units are: 



I. The Australian Fauna: Occurring all around the coasts 

 of Australia, and including the Aru Islands and southern New 

 Guinea (Papua). 



The Australian fauna, which is Httoral and sublittoral only, 

 is characterized by certain very distinctive, primitive and aber- 

 rant specific types. Three genera, Comakdella, Oligometrides 

 and Ptilo7netra are confined to it, while Zygometra and Petasometra 

 are here very highly developed. It is related to the Indo- 

 Pacific-Atlantic fauna, though it cannot be considered as a part 

 of it, or as a derivative from it. Of the several divisions of the 

 Indo-Pacific-Atlantic fauna it is most closely allied with the 

 Caribbean, and two of the three peculiar genera are, outside of 

 the Australian region, most closely related to Caribbean types, 

 Comatulella to the genus Coinactinia, and Oligoinetrides to the 

 genus Analcidometra; it is interesting to note that neither Comac- 

 tinia nor Analcidometra occur on the eastern shores of the Atlan- 



