BY E. J. TILLYARD. 291 



In concluding this paper, we may briefly review the position of the F'anor- 

 poid Orders as revealed to us in Upper Permian times by these fossils. At the 

 present day, the six main Orders fall into three groups of two each, viz. ' 



(a) Mecoptera and Diptera, characterised by simple Cui and dichotomic 

 branching of Rs. 



(b) Trichoptera and Lepidoptera, characterised by forked Cui and dicho- 

 tomic branching of Ks. 



(c) Megaloptera and Planiijennia, characterised by forked Cui and pec- 

 tinate branching of Es. 



Each of these groups is now seen to have been represented by Upper Per- 

 mian ancestors in Australia, 



(a) by true Mecoptera of the family Permochoristidae, and by the genus 

 Parahelmontia of the Parameeoptera.. 



(&) by the genus Belmontia of the Parameeoptera. 



(c) by the genus Permithone of the Planipennia. 



We are able, from this, to see that two Orders, the Mecoptera and Plani- 

 pennia, were already in existence in Australia in Upper Permian times. On 

 morphological grounds, we may also postulate the existence of true Megaloptera 

 somewhere in the world at the same period, though not necessarily in Australia. 



The history of the three more specialised Orders is now fairly plain. The 

 type represented by Parabelmontia gave origin, in the Triassic period, to the 

 main mass of the Paratriehoptera, from which the Diptera arose directly by re- 

 duction of the hindwing. The type represented by Belmontia gave origin, pro- 

 bably also in the Trias, to the common Trichoptero-Lepidopterous stem (almost 

 certainly far outside Australia), and the two Orders became differentiated either 

 in the Upper Trias or Lower Lias, the Lepidoptera remaining as an obscure 

 group of Homoneurous types until the rise of the Flowering Plants in the 

 Cretaceous brought with it the great development of the Heteroneura. 



If the whole of the Insecta Holometabola had a common origin, as I be- 

 lieve to be the case, then it follows that both the Coleoptera and the Hymenop- 

 tera must have been represented by primitive types in the Upper Permian, or 

 even earlier; since both these Orders are, morphologically, older in some respects 

 than the Panorpoid Orders. Consequently we should expect to find, though not 

 necessarily in Australia, primitive fossil beetles, allied probably to the Cupedidae, 

 and primitive fossil Tenthredinoid Hymenoptera, somewhere in the higher 

 Palaeozoic strata, in some part of the world. 



Cawthron Institute, 20.2.22. 



EXPLANATION OF PLATES XXXIII.-XXXIV. 

 Plate xxxiii. 

 Fig. L Archipanorpa (?) bairdae. n.sp. (x 7). 

 Fig. 2. Parabelmontia permiana, n.g. et sp. (x 5.8). 

 Fig. 3. Permithone belmontensis, n.g. et sp. (x 11.5). 



Plate xxxiv. 

 Fig. 4. Pincombea inirabilis, n.g. et sp. (x 15.6). 

 Fig. 5. Permochorista sinuata, n.sp. Holotype obverse, (x 11.5). 

 Fig. 6. Permochorista sinuata, n.sp. Paratype. (x 11.5)- 

 Fig. 7. Permochorista affinis, n.sp. (x 11.5). 



