134 DR. J. STEPHENSOX ON THE MORPHOLOGY, CLASSIFICATION", 



for the qiiick-moviug- Camivora ; in asking for land-bridges to 

 explain the distribution of the OJigoeha^tni we get much more than 

 we want. There can never have been a. land connection between 

 Australia and the great land mass to the north-west since 

 the Eocene. 



Still stronger is the cnse of the supposed connection between 

 India and Ne\^' Zealand. This is a necessity, according to 

 Michaelsen, in order to explain the occurrence of the Octo- 

 cha?tin9B in both lands ; and since the Octoclipetinas do not 

 occur in Australia, the bridge in this case avoided Australia. 

 ]S^ew Zealand does not even contain Marsupials ; yet Octochcetus, 

 the genus common to India and. jS'ew Zealand, is not a particularly 

 archaic genus, and its occurrence in both India and New Zealand 

 would, on Micliaelsen's view, have to i)e explained bv, presumabl}', 

 late or middle Tertiary land connections. But New Zealand is 

 an oceanic island, and probably has never been connected at any 

 time* with the larger land-masses, certainl}^ not in Tertiary 

 times. 



It is quite possible that similar objections might be brought 

 against che other land-bridges which have been postulated to 

 explain the existence of I'elated or identical genera of earthworms 

 in distant lands. I have specially mentioned the above because 

 it is so obvious, once attention has been drawn to it. The general 

 principle is that, ea,rthwoi-ms being a recent group, and requiring, 

 on the hypothesis of dis]:)ersal by land, counections of some 

 considerable permanence, other groups will have been able to 

 pass even more easily ; and the dispersal of earthworms by land- 

 bridges cannot be assumed unless there is a large degree of 

 similarity between other elements of the fauna, also. 



(4) Contributions towards a more satisfactortj Sohition. 



I trust that, in what follows, I shall not be considered to be 

 treating too lightly the claims of zoogeography to a hearing in 

 the discussion, of the problems of jDalseogeograph}-. As Michaelsen 



* Micliaelsen's time-sclieme cau be put together soinewliat as follows : — Tlie 

 oldest components of tlie Indian earthworm fauna date from the Upper Jurassic, 

 when India was connected broadly with lioth Angara and Australia; FhifeUiis and 

 JEegascoUdes wandered oft" into Angara, reaching western N. America in the later 

 Cretaceous. The chief part of the evolution took place in the Tertiarj-, the period 

 of the changing- land-bridges. In the Pliocene the now consolidated Indian 

 peninsula became connected on the W. or N.W. with lauds which had earlier 

 received their earthworms from Tropical Africa (Etidi chop aster). 



It will be seen that he puts the evolution of the group earlier than I do : but I do 

 not find an^'thing which invalidates the line of argument and general conclusions of 

 section 3a above, especiall.y that of the c[aite recent origin of the phyletically youngest 

 genera such as Megascolex. The word used hj' Michaelsen for the period of the 

 origin of the Indian Oligochaste fauna is "Malm," which corresponds (Ziegler, 

 Zool. Worterbuch) to the Upper Jura. PluteTlns and MegascoJides are supposed 

 to have then been in existence ; is there any other example of genera ot a variable'and 

 evolving group persisting since that period, especially genera, such as these, which 

 are connected by intermediate gradations not only with each other, hut with the 

 genera below and above them (J)iplot)'ema and Notoscole.r) , genera, that is, which 

 are still not sharply marked off" from their ancestors and descendants? 



