BY B. J. TILLYARD. 657 



• 



in those highly specialised types in which, after Rs has already 

 hecome simplified, further reduction leads to loss of branches 

 of M in the hindwing, the main stem of M is always obliterated. 

 Thus there are no known types within the existing Heteroneura 

 winch reproduce the condition of the hindwing of the Palaeon- 

 tinidae. 



The only conclusion that I can come to is that, although the 

 Palaeontinidae in which the hindwings have been preserved are 

 most certainly Heteroneurous, they do not belong to the line 

 of descent from which all our present-day Heteroneura are de- 

 rived. They must be considered as a separate offshoot from 

 the Homoneura, specialised by reduction of the branches of M 

 in the hindwing, and therefore having left no descendants at 

 the present day, and with no near relatives amongst present-day 

 Heteroneura. Handlirsch, therefore, in arguing in favour of a 

 relationship between these moths and the Limacodidae (2, p. 

 619), is not on firm ground, and his contention should be aban- 

 doned. 



We may now return to the study of the family, and consider 

 whether the venation of the forewings of those genera in which 

 the hindwings are preserved is sufficiently near to those in 

 which they are not, to justify us in accepting Handlirsch's 

 placing of them all as closely related types. 



The answer to this question must undoubtedly be in the affirm- 

 ative, for the following reasons : — 



(1) All the fossil types agree in having the branches of M in 

 the forewing occupj'ing the middle distal portion of the wing, 

 whereas the branches of Rs are crowded up anteriorly, as in the 

 higher Heteroneura of the present day. If we compare this 

 with the condition to be seen in the Hepialidae (Text-figs. 81, 

 82), we see how little justification there is for Meyrick's asser- 

 tion that "there is little doubt that it (i.e. Palaeontina oolitica) 

 belongs to the Hepialidae" (17) . The forewing of Palaeontina 

 is particularly close, in this and most other respects, to that of 

 Prolystra lithograpliica Opp. (Text-fig. 84, d), one of the forms 

 in which the hindwing is preserved. 



(2) As regards the condition of M 4> there is a gradation from 

 the oblique position, such as is found in the Hepialidae, through 

 the intermediate condition seen in Palaeontina, to the transverse 

 position seen in Limacodites (b) and Prolystra (d). Of all the 

 forms figured, the most archaic condition of M 4 is to be seen in 



