304 THE CANADIAN ENTOMOLOGIST. 



the tubercles are primary, not recently developed structures. Automeris 

 and Hemileuca must have separately acquired stinging spines and not 

 from " an initiatory existence in a common ancestor," as Mr. Grote puts 

 it, because there is no such condition in Aglia or Citheronia, which are 

 placed between them, nor any evidence that these genera are derived 

 from wart-bearing ancestors, a condition necessary for the multiple spine 

 formation. I regard these conditions as very improbable. 



If my larval tree is correct, Aglia has remained more nearly in the 

 primitive condition in regard to vein IV^ in that this vein is only slightly 

 moved toward IVi. Yet, it is somewhat moved, which weakens Mr. 

 Grote's point (see the fig. Die Sat., p. 19, fig. 8). Hemileuca has followed 

 out the same process separately ; derived from a common stem with 

 Automeris, vein IV2 has moved close to IVi separately from the Attacus 

 branch. Hemileuca is separately specialized in this respect. 



Some collateral evidence may help -to a conclusion. If these moths 

 be separated on the position of the wings in rest, the wings folded over the 

 back in the shape of a roof or with the upper faces together (as in butter- 

 flies) a tree results like the larval one. In the hind wings there are two 

 anal veins in Hemileuca and Citheronia, the rest have one. A tree con- 

 structed on these characters (which I believe to be as good as the one 

 selected by Mr. Grote) would be different from either. To reconcile i^ 

 with my tree, it must be supposed that the left-hand branch had lost one 

 anal vein, while Automeris on the right branch also lost it, but separately. 

 To reconcile it with Mr. Grote's tree, three separate losses of anal vein 

 must be supposed, viz., in Automeris, Aglia, and Attacus-Saturnia. My 

 view is here the simpler. 



As to the pectinations of the antenna? in the male, those of Hemileuca 

 only have simple branches ; in the female only Attacus-Saturnia have 

 them double. Mr. Grote may suppose that the original ancestor had 

 single pectinations in both sexes, retained in Hemileuca ; the right branch 

 acquired double pectinations in the male, while Attacus-Saturnia separately 

 acquired them in both sexes, which seems improbable. 



In my larval tree the ancestor must have already possessed double 

 pectinations in the male, which became transferred to the female also in 

 Attacus-Saturnia, but were lost by degeneration in the male of Hemileuca. 

 It is true that this supposition can also be applied to Mr. Grote's tree, so 

 that we are not greatly benefited by the consideration. Other characters 

 will have to be compared ; but this I will leave to Mr, Grote, with the 



