372 ENTOMOLOGICAL SOCIETY 



abdominal segment. This would seem to prove that wart forma 

 tion has taken place twice independently in the Noctuidae, and it 

 agrees with the indication of the moth structure. 



It will be remembered that Prof. Smith began his phyloge- 

 netic tree with the Erebiini and Homopyralini as one branch, 

 which terminates, and the Poaphilini, Catocalini, and Phoberini 

 as a second branch. All these have generalized larvae without 

 exception, so far as my material goes, though in a few species of 

 Catocalini and Phoberini, the terminal twigs, a slight tendency 

 toward specialization shows in that tubercle iv on the sixth ab 

 dominal segment has risen as high as the middle of the spiracle. 

 The Phoberini give rise to the rest of the groups in two main 

 stems. The first throws off the Eucalypterini and Cillini, of 

 which I only know one larva, which shows partial specialization, 

 being intermediate. Next the Acontiini and Erastriini ; of these 

 I have five larvae, two generalized, one intermediate, and two spe 

 cialized. Finally the branch ends in the Anomiini, Plusiini, 

 Stirriini, Eudryini, and Heliothini. All of these have general 

 ized larvae except the terminal Heliothini, which are specialized, 

 though not very strongly so. The other main stem from the 

 Phoberiini gives two branches, the Orthosiids and Hadenids. 

 The Orthosiids branch into the Mamestrini, of which I have 19 

 larvae, 16 specialized, 3 generalized; the Orthosiini, 8 special 

 ized, i generalized ; the Xylini, 5 generalized, 3 specialized. The 

 Orthosiini give rise again to theCosmiini andNoctuini, of which 

 I have 27 larvae, three only being generalized. The Orthosiids 

 are thus decidedly specialized with the exception of the Xylinini. 

 This group is given by Prof. Smith as a side branch from near 

 the base of the main stem. The Hadenids branch into the Cara- 

 drini with three generalized, one intermediate, and two specialized 

 species, and the Hadenini with three generalized and ten special 

 ized species before me, as one furcate branch. The other branch 

 gives in succession the Bryophilini and Acronyctini all fully spe 

 cialized, without exception, in my material. 



Thus we see a generally advancing tendency for the majority 

 of the species to become specialized as we advance up Prof. 

 Smith's phylogenetic tree, and in no case do the larvae essentially 

 contradict Prof. Smith's results. 



The following list will show just which genera and how many 

 species in each have been examined in the larva. G for general 

 ized, S for specialized. 



