TILLYARD. 161 



masses of the present day. In all stages of their life-history, Blepharoceridae 

 are dependent on the rushing water and spray of waterfalls, and are quite 

 unable to exist for more than a very short time without these. Hence their 

 distribution cannot have been brought about by sea or air carriage, but must 

 have taken place along definite land routes marked by the frequent occurrence 

 of running streams; and this, of course, indicates land of a mountainous nature. 

 The morphology of the family would indicate for it an origin in either Liassic 

 or Upper Jurassic times. Thus the presence of a species of Edivardsina on 

 Mount Kosciusko is best explained by the Antarctic Theory, which postulates 

 that, at some ancient time or other, but not necessarily at the same time, Aus- 

 tralia and Tasmania together, and New Zealand and Patagonia separately, were 

 all linked with the Antarctic Continent when it enjoyed a temperate climate, 

 and thus allowed of the passage of many types of plants and insects from one 

 to the other of these regions. If this theory be correct, it may reasonably 

 be expected that another species of Edivardsina, or of some closely allied new 

 genus, will be discovered in the high mountains_ of Tasmania. 



The other genus discovered on Mount Kosciusko is Neocurupira Lamb., 

 found only hitherto in New Zealand, and also related to Curupira from Brazil. 

 This genus also may be expected to occur in Tasmania. It is considerably more 

 highly specialised than Edivardsina. 



The genus Apistomyia Bigot, to which the new species found at Wentworth 

 Falls belongs, is represented by species found in Corsica, Cyprus, the Himalayas 

 and North Queensland, and will probably be found to exist in all suitable localities 

 right through the Oriental Region and across to New Guinea as well as down 

 the Eastern Coast of Australia. It is absent from New Zealand, Campbell's 

 Apistomyia harrisi, recently described, (Trans. N.Z. Inst., liii., 1921, p. 262), 

 being undoubtedly a true Peritheates, and this latter genus being so far unknown 

 outside New Zealand. It is more highly specialised than Neocurupira. Thus, as 

 we pass from South to North along the Eastern ranges of Australia, we meet 

 first with the more archaic types, and then with the more highly specialised. 

 This, again, suggests a Southern origin for the family in Australia. New Zea- 

 land, likewise, on present evidence, has only received types from the South; and 

 the comparative much greater abundance of these insects in the South Island 

 than in the North also bears this out. 



Bezzi has divided the family Blepharoceridae into four subfamilies, I have 

 not adopted these, and do not think it necessary to discuss them, as it seems to me 

 very clear that such divisions are very unwise in the present state of our know- 

 ledge, and that much more work needs to be done on the family before the main 

 lines of evolution within it can be made clear. Until that is done, preconceived 

 opinions of relationships of genera, based on their supposedly belonging to one 

 or other of Bezzi's subfamilies, only prevent a clear view of the field, and are 

 apt to bias or predispose one's judgment in the matter. What is perfectly clear 

 is that Edwardsina is easily the most archaic genus so far discovered, and that 

 the archetype of the family was not far removed from this type. Consequently, 

 all known genera can be derived from a type not unlike Edwardsina; but how 

 many separate lines of descent there are actually represented by living forms at 

 the present day, we cannot say for certain. A careful study of larval and pupal 

 characters may help us to solve this problem; but the principal characters for 

 each genus must always be drawn from the imago, as has hitherto been done 

 by all workers in the group. 



In the following Key to the Genera, I include the New Zealand genus Peri- 

 theates Lamb, as it seems likely that it too may yet be discovered in Australia or 

 Tasmania, 



