1 24 NATURAL SCIENCE. Feb., 



It should be mentioned that in the above list those species which 

 are peculiar to the islands in question are marked in italics, those 

 which are found elsewhere, as well as in those islands, in roman 

 character, while species incevtcB sedis are enclosed in brackets. 



The list is unfortunately by no means imposing, but it sums up 

 the facts at our disposal. I have, of course, not considered New 

 Caledonia as belonging to the category of oceanic islands ; and there 

 are some who will find fault with me for placing Mauritius and the 

 Seychelles in this list at all. Otherwise, I think that all the islands 

 which I include are true oceanic islands of either volcanic or coral 

 origin, which have never formed part of a pre-existing continent. It 

 is unfortunate, too, that so large a proportion of the species are 

 incerta sedis ; most of these were described by Kinberg in the Ofversigt 

 of the Swedish Academy for 1866, at a time when the structure of 

 this group of animals was very little known; his "Lumbricus" is 

 far from being coextensive with that genus as at present understood. 



The facts, so far as they enable any generalisations to be formed, 

 seem to indicate that earthworms are among those groups which 

 have the greatest difficulty in crossing the sea by the usual means of 

 transit open to such creatures. Very few exact experiments have 

 been made; but what we do know appears to indicate that salt 

 water is fatal to them. This statement cannot be made universally. 

 There are a few — very few — species which habitually live upon the 

 sea shore ; this is the case with Pontodrihis liiovalis of the Mediter- 

 ranean coast, and of the allied species P. berinudensis of the Bermudas, 

 Brazil, and Jamaica. Schmarda described, under the name of 

 Ponioscolex arenicola, at least three species from the shores of Jamaica, 

 of which one is the same as that subsequently described by Perrier 

 as JJrochceta hystrix, a very widely distributed form ; Pontodrihis bevmu- 

 densis was also confused under the same specific name, as also was 

 a third species which I have called Pontoscolex arenicola, preserving 

 Schmarda's original name. It has been also asserted that certain 

 Ceylonese earthworms are not killed by sea water. It will be 

 observed, however, that the facts contained in the list which I here 

 give argue strongly that it is only exceptionally that earthworms can 

 have crossed the sea to oceanic islands by the help of trees floating 

 along with currents, the only natural method that suggests itself. 



In remote oceanic islands, such as the Azores, the species are all 

 widely-distributed forms. Perichata indica, for example, is one of the 

 commonest species to be met with in gatherings of earthworms from 

 foreign parts, and it has also been found in greenhouses in this 

 country and in North America. This fact seems to show some 

 special adaptation on its part for accidental exportation by man. 

 P-recisely the same remarks may be made concerning Pontoscolex 

 ■ corethruvus and Eudrilus eugenics. The former species is almost the 

 most widely-distributed earthworm known. Specimens have been 

 recorded from various parts of South America, from the Malay 



