Oligochaefa.] SUBANTARCTIC ISLANDS OF NEW ZEALAND. 255 



certain Eiich\i:raeids in this way, it will not, I think, commend itself to the students 

 of other groups of animals and of jilants. The evidence in favour of a circumpolar 

 continent or shiftino; archi[ielai;i) is umch stroni;i'i- now than wlien T discussed the 

 matter in 1902. 



The subfamily Mega seal ecinnc is Australian in its origin, but representatives 

 of it have been recorded by myself from the North Island — species of Megnscolecidcs 

 (= Token. Benham) and SpencerieUa { = ])iporochaet(i) — and it is only very sparsely 

 represented by Ih'pnrochneta in the South Island and Stewart Island — so far, at least, 

 as our knowledge allows me to state. But, as I mentioned in 1904, while the common 

 earthworms of the South Island are Acanthodrilines, these are less numerous in 

 the North Island, where they are replaced by Megascolecines. It is therefore sur- 

 prising to find that the larger and commoner worms at the Auckland Islands belong 

 to this subfamily — viz., PluteUus (one species) and Diporochaetn (four species). 



Lumbricologists are agreed that the Megascolecines are derived from Acantho- 

 drilines — that, indeed, Notiodn/iis (s.l.) is the most archaic of the family, from which 

 Maoridrilus, Plngiochaetd, Rhododrilus, and Leptodrilus have been developed, in 

 somewhat that order. The Megascolecinae have, in the same way, been derived 

 from NotiodrUus through Phitellus, the most archaic of its subfamily ; and THporo- 

 cluK'ta is a descendant of it. (I am omitting reasons for these views, for they have 

 been fully dealt with by their author. Michaelsen, in a recent work, 1907, a.) 



The discovery of Phdellus on the Auckland Islands is, perhaps, the most astonish- 

 ing outcome of this expedition, so far as the earthworms are concerned. Character- 

 istically Australian, and chiefly Tasmanian and Victorian, its occurrence in this 

 southern outlier of New Zealand is difficult of explanation, for the genus has not yet 

 been recorded from the mainland nor from any of the islands in this area. It is 

 true that in 1904 I attributed a lacustrine species to this genus, but Michaelsen re- 

 gards it as belonging to Pontodrilus. In order to account for the presence here of 

 this genus, the following alternatives occur to me : (1.) It was introduced into the 

 islands at the time of the Enderby Settlement, from New South Wales. (2.) It 

 passed into and remained in the islands from New Zealand when this land had its 

 great extension southwards. (3.) It arrived from Tasmania, vin the Antarctic Con- 

 tinent. The last view seems to me the least probable. There is evidence that 

 Tasmania became separated from the Antarctic Continent before New Zealand did — 

 i.e., previous to the evolution of the Megascolecinae, at a time when NotiodrUus 

 and its immediate allies were the predominant genera of the family. Plutellus, indeed, 

 is a direct descendant of this genus, and no doubt appeared early. Again, if Plu- 

 tellus entered from the north, how is it that we do not find it on the mainland ? Its 

 descendant Diporochaeta is here, alongside the more archaic Acanthodrilids : and 

 it is clear that climatic and edaphic conditions of New Zealand are not antagonistic 

 to the Megascolecines. On the whole, I am inclined to take the view that the worm 

 was introduced by man (see below), for it is allied to cei'tain of the New S(Uith Wales 

 species, such as P. tuherculatus. 



The family Haplotaxidae is richly represented in the Southern Hetnisi)here, 

 for, with the exception of two or three species of Haplotaxis and one of Pelodrilus 

 in the Northern Hemisphere, the rest are known from the following places : In New 

 Zealand — two species of Jlaplotn.rix. one species of Pelodrilus .- in Western Austi'alia. — 



