BY R. J. TILLYAKD. 691 



% 



(Section w. The Venation of the Megaloptera. 

 ("Text-figs. 36a, 39-40, 44, 47a, 104-108.) 



Tins Order, as at present constituted, consists of a small num- 

 ber of archaic types not very closely related amongst themselves. 

 Probably none of their ancestors has ever been dominant at any 

 time, and this may well be the reason why, though the Order 

 contains chiefly forms with aquatic larva? and inhabiting the 

 banks of streams, the fossil record is exceedingly poor. 



Some authors, including Handlirsch (2), consider the terres- 

 trial Raphidioidea to belong to a separate Order, distinct from 

 the Sialoidea. Others again merge the Megaloptera with 

 the Planipennia into a single Order Neuroptera. 1 have, for 

 the present, taken the intermediate course of uniting the two 

 groups Raphidioidea and Sialoidea into a single Order Mega- 

 loptera, while keeping the Planipennia distinct. The important 

 point is not whether there are one, two or three Orders repre- 

 sented, but rather that we should keep clearly in mind the dif- 

 ferent lines of evolution indicated, whether we choose to class 

 them as belonging to one, two or three Orders. It will be seen 

 in the sequel that differences of opinion on this latter point 

 will not appreciably affect the definitions of the Archetypes in- 

 volved; and, as far as the venational scheme is concerned, at 

 any rate, the evidence will be found to favour only one conclu- 

 sion . 



The problem of determining the Archetype of the Megaloptera 

 is not an easy one. In order to grasp the problem fully, it will 

 be necessary to figure examples of all the different types known 

 m the Order. This I have done in Text-figs. 104-107. 



There are two Sub-orders, viz. the Sialoidea, consisting of 

 forms whose larva? are aquatic or semi-aquatic, and the Raphidi- 

 oidea, whose larva? are terrestrial. The latter are, in most 

 respects, rather highly specialised insects. But they cannot be 

 descended directly from the Sialoidea, because they all possess 

 the original condition of having Sc and R\ quite separate in 

 both wing's; whereas all the known Sialoidea have these veins 

 fused distally, as in the higher Planipennia. We may, for 

 the present, dismiss the specialisations shown by the Raphidio- 

 idea from eur minds, and only remember that this condition of a. 

 separate Sc and Rj has to lie present in any Archetype from 

 which they can possibly have been derived. 



