BICxELOW: MEDUSAE FROM THE MALDIVE ISLANDS. 249 



three is to bo explained on the ground that they all belong to the char- 

 acteristic "coral reef" type. Evidence that this cannot be considered a 

 truly representative tropical type is found in the fact that the Canaries 

 in the tropical Atlantic, whose physical characteristics, apart from the 

 temperature of the water, are very different from those of any of the 

 three other areas already considered, possess a Medusa fauna of mark- 

 edly different characters. 



As I have already stated, all of the Leptoline Hydroraedusae from the 

 Maldives, with one possible exception, are new. At the same time all 

 of the Trachylinae which belong to the families Geryonidae and Pegan- 

 thidae, whose members are well known to be local in their distribution, 

 are also new. In other words, all the "local" forms, with one possible 

 exception, are new, and the only species of Hydromedusae already known 

 are those distributed, or at least represented, by exceedingly close allies 

 throughout the tropical oceans of the globe. This same rule holds 

 good for the Discomedusae, Aurelia, and jSTausithoe, the Siphonophores 

 and Ctenophores. We reasonably expect to find traces of such a condi- 

 tion in almost any region. The striking thing in the Maldives is the 

 extent to which it is seen ; for not only do we find nearly all the local 

 forms new, but we find them separated from their nearest allies by very 

 considerable divergences which amount often nearly to genei'ic impor- 

 tance. The frequent occurrence in the Maldives of very aberrant species 

 in genera which until now have been very homogeneous is a striking 

 feature. The main conclusions which I wish to draw from these facts 

 are two : — first, the very large proportion of new forms among those 

 groups whose members are known to be of somewhat local distribution, 

 particularly the Leptolina, and the fact that none of the typical Atlantic 

 or Pacific Leptolina were found, points to the conclusion that, so far as 

 the Medusa fauna is concerned, the Maldives are an area of geographic 

 isolation. The very considerable degree of divergence from their near 

 allies shown by the new species, and the frequent occurrence of aberrant 

 members in otherwise very homogeneous genera, points to the second 

 important conclusion, that this condition of isolation has lasted for a 

 considerable period. 



The fact that all but one of the genera of Acalephs found in the 

 Maldives occur in the Atlantic, while only about two thirds of them are 

 known to occur in the Pacific ; and that while we found no typically 

 Pacific genus, we did take five genera not previously recorded, except 

 from the Atlantic, — seems to point to a closer connection with the Atlan- 

 tic than with the Pacific. This connection, if it exists, is of very great 



