no THE VARIATION OF ANIMALS IN NATURE 



to intercourse) that many species of marine Crustacea (Cope- 

 poda — Scott, 1909 ; Euphausiacea — Hansen, 191 1) are homo- 

 geneous throughout very extensive areas and pass practically 

 round the world within certain isothermal limits. It is perhaps 

 curious that there is no gradual regional differentiation of such 

 species and that such mutations as occur are so rapidly and 

 effectively extinguished or spread throughout the population. 



Doubtless many of these exceptions may be ultimately 

 explained by reference to differences of habit, etc., which so 

 far are unknown. In some cases this seems to be very unlikely. 

 The contrast between the Oligochaeta and the land Mollusca 

 is a case in point. We are indebted to the late Lt.-Col. J. 

 Stephenson, F.R.S., for pointing out many facts in connection 

 with the slight variability of earthworms. He informed us 

 that undoubtedly many species are ' peregrine ' and are 

 carried round the world either as cocoons or adults, probably 

 in agricultural and horticultural produce. Michaelsen {fide 

 Stephenson) also postulates the action of winds in dispersing 

 the cocoons, but Benham criticises this view. Peregrinal 

 species like Allolobophora caliginosa are remarkably constant and 

 exhibit very slight or no variation over an enormous range, 

 and it would seem that the means of intercourse must be fairly 

 regular if local differentiation is so easily effaced (cf. marine 

 Crustacea). But there are also many species of earthworms 

 which are not thus peregrine and have a more localised range, 

 and these are invariably homogeneous. Lt.-Col. Stephenson 

 did not think that these species are accidentally transported 

 from place to place. Moreover the means of transport either 

 of cocoons or of adults (human agency, birds, winds) should 

 be also similarly operative in the case of land snails. 



It remains to notice some theoretical considerations which 

 have a bearing on the interpretation of these facts. In the 

 first place we must emphasise the difference between small 

 local assemblages having a distinct statistical expression and 

 larger ' geographical ' groups. The greater average size of 

 Vertebrates must be of importance here, as it tends to involve 

 a wider range and less isolation. Small size, on the one hand, 

 facilitates accidental transport (Nematoda, Tardigrada), and 

 on the other makes for a homogeneous local population. In 

 the majority of cases the former influence seems to have been 

 paramount. 



