34 Transactions. 



anterior abdominal vein loses its direct connection with the sinus venosus, 

 and acquires a secondary one with the hepatic portal system." 



In the European frog Bombinator, which is closely allied to the New 

 Zealand frog Liopelma, Goette* had already described a similar course of 

 events, while Hochstetterf found in the larva of the salamander that the 

 two abdominal veins are at first separate for the greater part of their lengtli, 

 but unite near the liver to form a median vein which enters the left precaval 

 vein ; but with the absorption of the yolk this communication is lost and 

 a new one established with the hepatic portal system. 



The occasional persistence in Rana of either a right or a left connection 

 with the sinus venosus by way of a precaval has already been emphasized 

 by Woodland and O'Donoghue. 



That the present instance is not an exact recapitulation of the larval 

 condition is seen by the fact that the anterior end of the lateral vein on 

 each side is not connected with the sinus venosus but with the precaval, 

 which is apparently an earlier condition, as seen in the Elasmobranchs. 



Art. IV. — Some Earthworms from Stephen Island and D'Urville Island. 



By Professor W. B. Benham, D.Sc, F.R.S., Hutton Memorial Medallist. 



[Read before the Otugo Institute, 10th December, 1918 ; received by Editor, 30th 

 December, 1918 ; issued seixiratelij, 14th May, 1919.] 



During the present year Dr. J. Allan Thomson visited these islands in 

 Cook Strait, and amongst the animals collected by him were a few earth- 

 worms, which he was good enough to send to me for identification. 



No earthworms have hitherto been recorded from D'Urville Island, 

 though we are already acquainted with three species from Stephen Island. 

 Some were collected by Schauinsland during his visit in 1896-97, and others 

 by Thilenius ; and they have been described by Michaelsen (1) and by 

 Ude (2) respectively. 



These three worms are Ociochaetus thomasi Beddard, Maoridrilus tetra- 

 gonurus Michaelsen, and Dinodrilus gracilis Ude. The first of these species 

 is a well-known and widely distributed South Island form, which was one of 

 the earliest of the New Zealand worms to be accurately described (3) ; the 

 other two species are so far confined to Stephen Island, though they belong 

 to genera which are known in the South Island and the North Island. 



The present collection contains two out of these three species, and in 

 addition two new species of Maoridrilus, one from each of the islands, 

 together with a new species of Perieodrilus, a genus hitherto known only 

 from the mountains of the west coast of the South Island. So, as one 

 would expect, the Oligocliaet fauna of the two islands is mainly derived 

 from the South. 



* A. Goette, Entwickel der Unke, 1875. 



f F. HocHSTETTER, Morphol. Jahrbuch, vol. 21, p. 19, 1894. 



