MARINE MOLLUSCA OF THE KERMADEC ISLANDS 501 



Cosa costata (Bernard), Hoschstetteria meleagrima Bernard, Mytilus 

 canaliculatus Martyn, and Modiolaria impacta (Hermann). 



Upon close examination this already weak link proves weaker still. 

 Poirieria was recorded by the Challenger from 1,100 metres off the 

 Kermadecs and these shells have not been critically re-examined. My- 

 tilus was recorded by Suter but no specimens have been seen in quite 

 extensive collections. Such a species would undoubtedly be well re- 

 presented in beach collections if it were present. The Monia, Modio- 

 laria, Cosa, and Hochstetteris are all doubtfully the same as the New 

 Zealand forms. The Xenophalium occurs very rarely in northern New 

 Zealand waters, and Neothais smithi, though abundant at the Kermadecs 

 and occurring at Lord Howe and Norfolk, is but a rare straggler in 

 New Zealand waters. 



From the above it appears certain that molluscan dispersal move- 

 ments are from the Kermadecs to New Zealand, at least within recent 

 times. 



A number of the genera occurring at the Kermadecs are otherwise 

 endemic to New Zealand e.g. Eudoxochiton, Onithochiton, Maorichi- 

 ton, Austronoba, Haurakiopsis and Pinnoctopus (?) . If these genera 

 were originally derived from New Zealand it must have been at some 

 time in the Tertiary, since in all cases the forms at the Kermadecs have 

 differentiated specifically. In all these cases there is no palaeontological 

 evidence available to show whether dispersal took place in the direction 

 mentioned or from the Kermadecs to New Zealand, 



It thus becomes apparent that the marine molluscan fauna of the 

 Kermadec Islands has little relationship to New Zealand bvit has been 

 derived from a number of other sources. The geological history of the 

 group and their comparative isolation have been long enough for a 

 large number of endemic forms to evolve. Some of these are large and 

 characteristic shells e.g. Scutellastra, Tectus and the species of Cellana, 

 and Siphonaria. Indo-Pacific and Australian elements are well repre- 

 sented and a number of forms are shared with Lord Howe and Nor- 

 folk Islands. It is difficult to assign the gi^oup to any zoogeographic 

 region and a similar difficulty occurs with the neighbouring Lord Howe 

 and Norfolk Islands. The marine fauna of these three marginal areas 

 represents an interplay of factors such as geological age, distance from 

 neighbouring land, the direction of surface water movements and larval 

 motility and duration. 



Iredale and Allan (1940) suppose that the very evident relation- 

 ships between the marine mollusca of Lord Howe and New Caledonia 

 date back to a period when there was a land connection between the 

 two. The presence of Placostylus in both areas is fairly conclusive 



