678 



SUBANTAKCTIC ISLANDS OF NEW ZEALAND. 



[Land Nemerteans. 



characterized by the larger number of nerves. The barrier between the two groups 

 is further broken down by the possession by G. novae-zealandiae of numbers (13 

 and 17) characteristic of both groups. 



Let us now turn to the question of the number of reserve stylet-pouches. In 

 the table in Punnett's paper G. australiensis is stated to possess 2 and novae- 

 zealandiae 4. Now, Dendy, in his monograph on G. australiensis, says (p. 98), 

 "they are not definitely two in number, as appears to be the case in G. chalicophora, 

 but their number varies ; I have counted as many as five in a single specimen " ; 

 and, further, with regard to G. novae-zealandiae, " The calcareous stylets have the 

 usual form, and I detected four reserve sacs in the usual position. Probably, 

 however, this number is not constant." 



The explanation of the apparent contradiction between Coe's and Punnett's 

 table and Dendy's original account is probably of the following kind. 



I have been able to examine the reserve stylet-pouches in six specimens of 

 G. australiensis, and find them to vary in number between 2 and 5 ; but that even 

 when there are more than 2 the pouches are always arranged in two groups which are 

 vis-a-vis to one another, and in which the component individual sacs are closely 

 apposed to one another. The details of the association of these sacs in the two 

 groups, which we may call x and y, in the six specimens (a to /) are as follows : — 





Group X. 



Group y. 



Total 



a . . 



.. 1 



2 



3 



h .. 



2 



3 



5 



c . . 



.. 2 



3 



5 



d .. 



..2 



3 



5 



e . . 



.. 1 



1 



2 



/ .. 



.. 2 



3 



5 



It is therefore only true that G. australiensis has 2 reserve stylet-pouches in the 

 sense that it sometimes actually has only 2, as in e ; and that when it has more they 

 are always, so far as the present evidence goes, associated in two groups. In the 

 only specimen, however, of G. novae-zealandiae which I have which shows the dis- 

 position of these sacs the four sacs are not associated in groups, but are, as in 

 G. spirosfermia, at the four points of the compass, as seen in transverse section. 



It is possible that the disposition of the reserve stylet-pouches may afford a 

 further distinguishing character between the two groups. But it is a noteworthy 

 fact that the only two species of the other group which agree with the agricola- 

 chalicophora group in having their reserve sacs only 2 in number or arranged in 

 two groups are — (i) G. australiensis, which bridges the gap between the two groups 

 by the number of its proboscideal nerves ; and (ii) G. graffi, which possesses the 

 greatest number of proboscideal nerves in the whole genus — namely, 23. 



COMPARISON BETWEEN THE NEW SPECIES OF THE GENUS GEONEMERTES AND 



THOSE ALREADY KNOWN. 



G. spirospermia differs from any other known species of land nemertean in the 

 pattern of its pigmentation. There is no other species with two stripes. The pos- 

 session of spiral spermatozoa is also a character peculiar to it. It agrees with the 



