68 ILLINOIS BIOLOGICAL MONOGRAPIIS [310 



Catocala ? vidua presents the opposite situation, where we have an 

 arl)oreal form whose curve turns upward slightly in the last unit. Possibly 

 the larva of this species enters the soil to pupate, altho the members of this 

 genus typically spin cocoons above the ground. The epicranial stem of the 

 arboreal Heterocampa biUneata (Notodontidae) undergoes a marked 

 secondary shortening, showing that this condition in families other than the 

 Noctuidae is not necessarily associated with the subterranean mode of life. 

 In spite of the marked turning upward in the curve for this species, the stem 

 of the last instar is much longer than in larvae which enter the soil. The 

 peculiar curve of the bag-worm bears some relation, perhaps, to its unusual 

 feeding habit. A much more extensive postembryological study of this 

 structure must be made, embracing many families of lepidopterous larvae, 

 before we can hope to understand the significance of these changes. 



Having discussed the postembryology of the epicranial stem and its 

 biological significance, it now remains for us to consider the phylogenetic 

 evidence which this study may afford. A certain degree of correlation 

 between the types of curves in Plate I and taxonomic groups can be 

 observed. Attention has already been directed to the peculiar type of 

 curve presented solely by the two representatives of the Phytometrinae 

 examined. Whether or not this type is characteristic for the entire sub- 

 family we cannot state. When we consider the pronounced uniformity of 

 the larvae of this group, however, it seems fairly probable that this is so. 

 The species of Catocala represented exhibit an unusually great increase 

 in the length of the epicranial stem during the first two periods. Species 

 of the same genera have curves similar in position and shape, except where 

 the secondary turning upward has interfered. This process, being asso- 

 ciated with the subterranean habit, which often differs in closely related 

 species, cannot be relied upon as an indication of phylogenetic relationship. 

 With curves which turn upward the primary portions only can be safely 

 compared from a taxonomic point of view. For instance, Cirphis unipunda 

 presents a typical non-subterranean type of curve, whereas those of 

 phragmilidicola and pseudargyria, which are very closely related to this 

 species, are of the incipient subterranean type. The difference in the later 

 postembryonic development of this suture in A gratis c-nigrum and ctan- 

 destina has already been referred to. A comprehensive series of such 

 curves would unquestionably afford valuable phylogenetic information. 



All of the species examined were established earlier than the first 

 phylogenetic period represented in Plate I. In other words, none of the 

 curves of closely related species have started from the same point in the first 

 unit. A more extensive series might very possibly discover species so 

 recent that their curves would unite in a common line in the first one or two 

 units. 



