94 Transactions. 



the result of a colonization across shallow water, or from a land close at 

 hand, we should have expected to find a compact assembly, and not the 

 more or less scattered fragments that are actually found. 



Apart from the Australian element, the marine mollusca possess no 

 characters suggestive of recent accessions from other faunal regions, for 

 the Antarctic element is very largely a relic from older geological times. 



The terrestrial and fresh-water mollusca require for their dispersal a 

 close approach of lands, if not an actual land bridge. It is true that 

 occasionally marine currents may bring some species to oceanic islands, 

 but if these were introduced in some far-off period it would be extremely 

 difficult, if not impossible, to be certain as to their origin. 



Of the land mollusca which have been grouped under the Flammulinidae, 

 Endodonta and Laoma comprise by far the greater number of our snails. 

 They are ancient inhabitants and a primitive race. The geographic range 

 of the group is almost world-wide, but palaeontology so far has found very 

 little record of them. Pilsbry* remarks that " the Carboniferous of Nova 

 Scotia has afforded a small helicoid which in form and sculpture can only 

 be compared with such Endodontidae as Pyramidula or C7;aropa." In 

 support of our belief in the great antiquity of these helicoids, it may be 

 pointed out that a number of genera appear to have developed in our area 

 and are restricted to it, and that no New Zealand species has been recorded 

 from any other fauna! region. Of other groups, the Athoracophoridae may 

 perhaps have been developed in our region. The Rhytidae, which include 

 our large carnivorous snails, are of very ancient lineage. The operculate 

 group is in great measure peculiar to our fauna. Hedleyf remarks on this 

 when he discussed the relation of the fauna and flora of Australia to that 

 of New Zealand. Partly to account for the dispersal of Placostylus Hedley 

 constructed his Melanesian Plateau.! If that land area or archipelago be 

 granted a great antiquity it would appear to provide all the necessary 

 communications even for the most primitive forms. 



The groups of fresh-water mollusca are in perfect accord with the land 

 mollusca. Gundlackia, Pofamopyrgus, Lymnaea, Jsidora, and Melanopsis 

 are all of ancient lineage, while Diplodon is recorded from our Cretaceous 

 beds at the Malvern Hills. Hedley ,§ writing on the surviving refugees of 

 ancient Antarctic life, discusses many interesting problems of distribution. 

 He regards their advent as taking place by circuitous routes at wide intervals 

 of time, and thinks that they are of great antiquity. 



In our younger rocks the percentage of Recent species is very high, 

 and in making comparisons between these we prefer to use the names of 

 definite horizons where the species have actually been collected, rather 

 than period-names, which may often involve incorrect correlation. 



For the first of these we select the blue clays and sands of the coastal 

 cliffs near Castlecliff. The fauna of these beds differs from the Recent fauna 

 only in the presence of Ataxocerithium, and in the absence of a few obscure 

 genera and a few groups whose habitat is between tide-level or in very 

 shallow water. The total of extinct species is here not more than 7 per 

 cent. 



In the Kai Iwi beds, three-quarters of a mile south of the stream, the 

 fauna presents no distinctive features from that at Castlecliff, except 



* Manual of Conchology, vol. 9, p. xxxix, 1894. 



t Natural Science, vol. 3, p. 189, 1893. 



%Proc. Linn. Soc. N.S.W., vol. 7, ser. 2, pp. 337-39, 1893. 



§Proc. Roy. Soc, N.S.W., vol. 29, pp. 278-86, 1895. 



