1898] GEOGRA PEICAL DISTRIB UTION OF A CT INI ARIA 103 



pinnatificl forms in the Australian R. howesii ; or even with the 

 still further exaggerated condition met with in Gramhradis arabica 

 of the Red Sea, and in SarcopMcmthus sertus from New Ireland. 

 The vesicle-like nodulated tentacles of the West Indian genus 

 Actinoporus may in a like manner probably be regarded as a stage 

 towards the complication found in Cryptodcndron of the Indian 

 Ocean. It may be further noted that the majority of the Discosomae 

 of the Australian area are more gigantic than those of the Carib- 

 bean, and present tentacular differences. 



From these examples it is manifest that the conditions favour- 

 able for the production of Actinian complexity of form have been 

 better realized in the Australian and Red Seas than in Antillean 

 waters. 



Of the Zoanthidae, Haddon has remarked that, " so far as our 

 knowledge at present extends, the Macrocneminae \_Epizoanthus, 

 Parazoanthusj alone are represented in the North Atlantic, although 

 they are world wide in distribution." ^ This is in strong contrast 

 with the Caribbean area, where, excepting Sphenopus, all the 

 genera of the other sub-family, the Brachycneminae, — Zoanthus, 

 Isaurus, Geinmaria, and Palythoa — occur in the greatest abundance, 

 tlie first and last especially covering enormous patches of rock in 

 shallow water. Similarly in the Pacific and Indian Oceans the 

 Zoanthidae appear to be numerously represented, often by examples 

 not greatly differing from the Western species." 



From his examination of the Bahaman forms Professor 

 McMurrich was struck with the resemblance which the Actini- 

 arian fauna of those islands presents to that of the Pacific, and 

 its decided difference from that of the eastern coast of America, and, 

 as shown above, the Bahaman species closely correspond with those 

 of the more southern West Indian Islands. So far as the 

 Actiniaria are concerned McMurrich ('89, p. 69) defines two 

 great areas of distribution : " the Indo-Pacific, including the Indian 

 and Pacific Oceans and the seas connected with them, such as the 

 lied Sea ; and the Atlantic, including in this the Mediterranean." 

 His important conclusion is, however, that the Caribbean region of 

 the Atlantic should be separated from tlie Atlantic region and 

 united with the Indo-Pacific. Subsequent investigations carried out 

 in other areas, particularly those at Torres Straits and on the Barrier 

 Reef of Australia, serve but to emphasise this twofold division. But 

 to my mind, a further distinction must be made between the Caribbeo- 

 Pacific area and the Indo-Pacific, mainly in regard to the genera 



^ A Gemmaria has since been obtained from the Canary Islands ('96). 



'■* It is indeed difficnlt to distinguish by any external or anatomical character Isfmmii 

 asymmctricus, H. and .S., of Torres Straits from the West Indian /. tubcrculahifi. 

 Duch. The close similarity of the species of Zoanthus and Palythoa is also well 

 known. 



