122 Transactions. — Zoology. 



this colony, and it may therefore help to swell the already 

 long list of shells erroneously ascribed to New Zealand. 



4. Gundlachia, sp. Plate XIV., figs. 1-5. 



About two years ago I found in the Eiver Avon, below the 

 outflow of the Horseshoe Lake, a minute ancyliform shell, 

 which I could not separate from AncyUis woodsii, Johnston 

 (figs. 1, 2), from Tasmania. I then found only empty shells, 

 but further collecting furnished a good number of them alive. 

 On examining the shells I found, to my great surprise, that in 

 several of them the base was more or less closed by a septum, 

 as it is found in the Tasmanian Gundlachia (fig. 3). On con- 

 sulting Johnston's papers on the Tasmanian fresh-water shells 

 I found the statement that his Ancylus woodsii has " the 

 animal and teeth almost similar to Gundlachia petterdi" (!), 

 and that in the young state the shell of Gundlachia resembles 

 the common Ancylus. I compared the dentition of our shell 

 with that of a Gundlachia from Ohio, and there was almost no 

 difference ; therefore the shell from the Eiver Avon must be 

 considered as a Gundlachia. Figs. 4 and 5 show the form of 

 jaw and teeth. The shell, on attaining its full development, 

 will, no doubt, resemble the Tasmanian G. petterdi, but I 

 have not yet succeeded in finding it. According to my own 

 observation, and information received from Tasmania, Gund- 

 lachia seems to attain but seldom its full development, but 

 grows cind dies mostly in its ancyliform shell, without even 

 attempting to form a septum. This is shown by the fact that 

 here, as well as in Tasmaiiia, Ancyhcs ivoodsiiis abundant, and 

 Gimdlachia is rare in the same locality. 



Professor Hutton told me that this Gundlachia in the 

 Eiver Avon might possibly have been introduced from Tas- 

 mania on aquatic plants used for packing trout-ova. This 

 may be, but I rather doubt it, for the following reason : I have 

 not found yet the shell in question from the outflow of 

 Horseshoe Lake upicards to the fish-hatching establishment, 

 a distance of several miles, but only in that outflow and down- 

 wards from its disemboguement in the Eiver Avon. Very 

 likely it is living in that lake, and was brought down to the 

 river when the canal was cleared from w^eeds. The lake is not 

 easy of access, and I have not had an opportunity of exploring 

 it. The question whether it is an introduced form or not 

 can only be settled with certainty when it is found in a 

 locality where the above-mentioned mode of introduction is 

 out of question. 



The fact of Gundlachia occurring in New Zealand would 

 not be astonishing at all, for we have, besides some genera of 

 fresh-water shells, of the land-shells the sections Flammulina, 

 Gerontia, Fhacussa, Allodiscus, Thalassohelix, Fhrixgnathus, 



