no NATURAL SCIENCE. Aug., 



the truth of the statement that in all these cases the foreign species 

 are identical with those found in Europe and North America. Were 

 the forms in question really indigenous, different species, or at the 

 very least, varieties might be fairly expected to occur ; but they do 

 not occur. Any difficulties in the existing distribution of animals are 

 apt to be explained in one or two ways ; either a continent is evoked 

 from the vasty deep for this occasion only, or tree trunks and floating 

 debris are pressed into the service. As has been already mentioned, 

 earthworms are killed by salt water ; so that we must assume a former 

 land connection to explain their presence on both sides of a strait of 

 whatever width. It is a most remarkable fact that works dealing with 

 geographical distribution and the geological arguments to be derived 

 from a study of animal distribution have not made use of this im- 

 portant material ; the only paper known to us which does make use 

 of this group is by Professor Spencer on the fauna of Tasmania (Aus- 

 tralasian Assoc. Adv. Science, Presidential Address, Section D, 1892). 

 Other terrestrial invertebrates are doubtless of use ; but up to the 

 present, hardly anything except insects and land shells have been 

 studied from this point of view. But in neither of these groups are 

 the anatomical relations between the genera and species so clear as 

 might be wished. This partly destroys their value, as does also the 

 fact that they possess facilities for migration, passive or active, not 

 possessed by the earthworm. Sealed up in its shell by the temporary 

 operculum, a snail is capable of extended travel ; butterflies can, as 

 we know, fly for long distances ; the earthworm has none of these 

 advantages; the cocoons might, it is true, be conveyed on the feet of 

 birds, but they are often below the soil, sometimes very far down ; 

 this mode of transit has very possibly occurred in the case of the 

 aquatic worms allied to the earthworm, which deposit their cocoons at 

 the margins of streams and rivers frequented by birds ; in this way 

 we may possibly account for the wide prevalence of the common "red- 

 worm " of our rivers (Tubifex rivulorum). 



Applying these facts to the particular case in question, it seems 

 to us that the former existence of a southern continent is rendered at 

 least very probable. The earthworm fauna of New Zealand, which 

 is fairly well known at present, is characterised by the presence of the 

 genus Acanthodrilus, this genus and its immediate allies forming 

 nearly the whole of the indigenous earthworm fauna of the country. 

 Australia, on the other hand, is characterised by quite a different 

 family, the Cryptodrilidae, comprising three or four genera, perhaps 

 more ; A canthodrilus is represented by only two species, which are 

 found upon that side which is believed to have had, at a remote 

 period, a connection with New Zealand. The Australian family is, 

 on the other hand, only represented by two or three species in New 

 Zealand. The genus Acanthodrilus (using the genus in its widest 

 sense) is also characteristic of Patagonia and South Georgia (so far 

 as is known), and the more southern parts of South America. The 



