140 EARTHWORMS AND THEIR ALLIES [ch. 



their lack of facilities for migration, other than unas- 

 sisted locomotion, points which have been dealt with 

 earlier, it is difficult to explain their range in the 

 antarctic hemisphere on other grounds. The very 

 fact that the actual earthworm fauna of New Zealand 

 has led us on the whole to assign it to the Indo- 

 Australian regions shows the inherent uselessness of 

 the current view of zoogeography. For were we to 

 leave the matter here the relationship of New Zealand 

 to the regions of the world which lie to the south of 

 it would not be apparent. However, here as in so 

 many cases there is an antagonism between cut and 

 dried systems and the indications of evolution. 



This assumed existence of a former antarctic 

 continent which connected Southern Africa and 

 Southern America as well as various islands has 

 perhaps a further justification in the distribution of 

 the Geoscolecidae. This family is divisible into two 

 well-marked sub-families of which one as has already 

 been mentioned is limited to South America and 

 another practically to Africa (the exceptions being 

 species of the largely aquatic Glyphidrilus), while 

 a third sub-family the Criodrilinae is more widely 

 distributed again in accordance, one may perhaps 

 assume, with its largely aquatic mode of life. It is 

 also conceivable that the genus Dichog aster is 

 another example pointing the same way. The argu- 

 ments for regarding this genus as an indigene of the 



