318 THE GEOLOGICAL RECORD [CH. 



present to confess. I would here refer to two disturbing questions 

 that inevitably face us. Firstly, what was the true relationship 

 between the Protodonata, as we take farewell of them in the 

 Permian, and the abundant and varied Mesozoic Odonate fauna 

 which we meet with in the Lias? Somewhere between these 

 two points, in late Permian times may-be, or in the Triassic, the 

 "old order" changed, "yielding place to new." To-day the veil 

 is not yet fully lifted, and we can only partly grasp the nature 

 of the profound changes by which the beautiful Odonate wing, 

 with its nodus, its pterostigma, and, in the Anisoptera, its intricate 

 triangle and radial-sector formations, came into being. For 

 European workers the problem is not to be solved, since the 

 Permian and Trias in that region offer little scope for such dis- 

 coveries. We may turn with some hope to the rich coal-measures 

 of New South Wales, and to the Mesozoic shales and later coal- 

 measures overlying them in that State and in Queensland. There, 

 bearing in mind that, all along, Australia has probably lagged 

 somewhat behind Palaearctic in the types of her insect fauna, 

 we may still hope to find the solution of this interesting problem. 



Secondly, what is the history of the Libellulinae, that fascinating 

 dominant group of to-day ? Its origin is lost in obscurity. Neither 

 palaeontology, ontogeny, nor comparative morphology has so far 

 been able to enlighten us. The absence of Libellulinae from 

 Solenhofen, and again from Florissant, may possibly be only due 

 to the nature of these deposits. We need to discover low-level 

 lacustrine deposits for the solution of this problem. We know 

 only of a single generalized species of Corduline from the Oligocene. 

 On the whole, it seems likely that the Libellulinae are a later 

 Tertiary offshoot of an older Corduline stem. They may have 

 arisen in Oligocene or early Miocene times, spread and increased 

 with great rapidity, and thus quickly become dominant over the 

 older groups of Anisoptera. 



On the other hand, let us look at the positive results of our 

 record. It tells us for certain that the Gomphinae, Petalurinae, 

 Cordulegastrinae and Epallaginae were well established in Jurassic 

 times, and differed little from these groups as we know them to-day, 

 except in a tendency to the possession of useless excess cross-veins. 

 Both ontogeny and comparative morphology agree with the 



