70 Papers from the Department of Marine Biology. 
III. CONCLUSIONS. 
As already stated, the data at present available are too fragmentary 
to warrant any reliable deductions. We can as yet scarcely guess at 
the origin of the littoral fauna of the West Indian Islands, but certain 
things are suggested by this study which may be mentioned as requir- 
ing further consideration: 
1. There is no very close relationship with the Mediterranean fauna. 
Of the 55 genera concerned, only 25 occur in the Mediterranean, and 
only one of these is as yet unknown on the western coast of America. 
2. There is a notable resemblance to the fauna of the western coast 
of tropical America, four-fifths of the genera (44) being known to occur 
there. And in many genera specific differences between the West 
Indian and West Coast forms are very slight. 
3. The fauna of the Bermudas is practically all derived from the West 
Indies, and so recently that no endemic species have as yet arisen. 
The only endemic species in Bermuda is Leptosynapta acanthia, which 
is probably derived from one of the northern members of the genus. 
4. The fauna of Tobago unquestionably contains a southern element 
derived from the Brazilian coast. 
5. If we assume that the genus Arbacia arose on the western coast 
of America, the present distribution of the genus and of the species 
punctulata can be explained as follows: The Caribbean Sea was at 
one time an eastward extension, a narrow-mouthed gulf of the eastern 
Pacific ocean, formed after Arbacia was well distributed north and 
south of the present isthmus of Panama. The present species punc- 
tulata entered this gulf on both shores and followed them eastward. 
After the closing of the gulf and the formation of the Lesser Antilles, 
the species being exclusively littoral spread, not in all directions, but 
only northward along the Mexican and United States coasts and 
eastward along the South American coast to Trinidad and Tobago. 
Later the genus passed southward to Brazil and eastward to Africa 
and the Mediterranean in the form of lirula, which may well have 
been derived from punctulata. Along the coast of the United States, 
the local conditions have not been favorable to developing a new 
species. From Florida, punctulata has crossed over to Cuba and 
perhaps extended along the north shore of that island to Haiti. 
According to this theory, Arbacia does not and never has occurred in 
the Lesser Antilles nor in Jamaica and probably does not occur on 
the south side of Cuba or in Porto Rico. It may be found in the 
Bahamas. It certainly ought to occur in favorable places along the 
north coast of South America. Until it can be shown to occur there, 
the hypothesis here proposed with reference to Arbacia lacks adequate 
foundation. But the distribution of Hemipholis, Echinaster, Moira, 
Meoma, Thyone, Stichopus, and similar genera, as shown in the table 
on pages 71-73, gives some support to the theory and warrants its 
consideration in interpreting the West Indian fauna. 
