20 MR. R. J. TILLTARD : LIFE-HISTORTES AND 



SO, 710 special significance can he attached to its present occurrence in tioo so" 

 widely/ separated localities ; for these remain simply as the last refuges of a 

 dying group. 



B. They may be the remnants of a group that developed in Antarctica at a 

 time when that region enjoyed a temperate climate, and were, later on, driven 

 out along two separate land-connections — the one leading to the mountains of 

 Chili, the other to those of Eastern Australia, If this be so, the Petcdiini 

 afford direct evidence of a former land-connection, vid Antarctica, between 

 Australia and South America. 



Some light may be thrown upon this question by studying the present 

 distribution of another archaic remnant of the ^Escbnidse, viz. {h^Petalurince. 

 Seven speeies are known, referable to four genera — just exactly the same 

 number of species and genera as in the case of the Petaliini. But the forms 

 are not so closely allied to one another, and their distribution is more widely 

 spread. Three species (genus Petalura) occur in Eastern Australia ; one 

 species (genus Uropetala) in New Zealand ; one species (genus Phenes) in 

 Chili ; and two species (genus Tachopteryx) in North America. 



In a case such as this, supposition A is the only possible explanation. We 

 could scarcely argue that the distribution of the Petalurince really offers- any 

 evidence of a land-conection between Australia and the New World. If in 

 course of time the genera Tachopteryx and Uropetala became extinct, the 

 Petalurince would then offer to us a distribution very similar to that of the 

 Petaliini at the present day. 



Against this line of argument, and in favour of supposition B, we can only 

 offer very little evidence. The strong point in favour of B seems to me to 

 be the exceedingly close agreement between the Chilian and Australian 

 forms of Petaliini. In the case of the Petalurince the typical expanded! 

 superior appendages of the male Petalura and Uropetala (Australian group); 

 are not found in the American genera. There is also a considerable 

 difference in venational and other details. But in the Petaliini all| forms- 

 seem to have preserved not only the wonderful wing-spots, but also the 

 peculiar trifid form of inferior appendage of the male.. So striking are these 

 characters, and so closely similar is the Australian species tO' its Chilian, 

 relatives, that one is much tempted to explain them as two separateflines of 

 development from a single Antarctic genus, which was^ driven out from that 

 region along two distinct land-cannections, and has since diverged but 

 slightly from its original form. 



There is now incontrovertible evidence that the climate of Antarctica was- 

 once temperate, but we cannot yet be certain how late it remained so. Fossil 

 Odonata of the Jurassic and succeeding periods are fairly rich in forms not 

 far removed from iiie Petaliini (e. g., Cymatophlehia), so that it is quite- 

 possible that the Petaliini themselves became established about Cretaceous- 

 times. This woull give plenty of time for the required segregation of the: 



