8 LYCtENA IL, III. 



larva? willow leaves, which they ate. At Yonkers, N. Y., Dr. Howe saw females 

 of Lucia ovipositing on Dogwood ia April and May, 1878. So far as I am aware, 

 this comprises all that has been published respecting the preparatory stages of 

 any of these forms outside of my own observations. 



The inter-relationship of the forms is complicated, but will be found nearly as 

 follows : — 



1. The chrysalids from the late larvte, which feed on Actinomeris, hibernate 

 and produce Violacea in spring. 



2. The chrysalids from Violacea, in spring, in part disclose Neglecta, in June 

 following, and are the parents of that brood, Ijut most of them hibernate, and 

 produce Violacea the next year. So far, I have not succeeded in carrying Vio- 

 lacea chrysalids through the winter. They have died late in the fall or early in 

 the winter, either from being kept too dry in the house, or from mould when I 

 have endeavored to keep them damp. But two which were found dead 27th 

 November, on having the wing cases removed, showed the full colors of Neglecta, 

 1^1?. I have at the date of this writing, 14th January, 1884, one chrysalis of 

 last spring which is alive. These facts show sufficiently the tendency of part of 

 the chrysalids to hibernate. In no other way than by hibernation can the mul- 

 titudes of the butterflies of this form in spring be accounted for, as the few 

 larva? and fewer chrysalids of the fall can produce but the merest fraction of that 

 flight. 



3. The chrysalids from the May generation, or Pseudargiolus, probably pro- 

 duce butterflies in small numbers in July and later, after the June Neglecta have 

 passed away, but most of them hibernate, and give Pseudargiolus the following 

 May, or earlier. I do not know that I have had a butterfly emerge the same 

 season from a chrysalis of this form, as I found in 1883 that previous observa- 

 tions were imperfect, because till this year I had confounded the larvo3 of Neg- 

 lecta feeding on Cimicif uga with those of Pseudargiolus on same plant ; that 

 is, the late larvaj with the early ones. But on 27th November, 1883, out of 

 twenty-seven chrysalids of Pseudargiolus, which formed between 20th June and 

 8th July, six were dead, and the wings of 1$ 2.9 showed full color of Neglecta; 

 two were partially colored, one not at all. Five then died when about to issue 

 from chrysalis, and this is proof that part of the chrysalids of this form give but- 

 terflies the same season. The remaining twenty are alive on 14th January, 1884. 



Out of twenty-five chrysalids from Neglecta, formed between 1st and 22d July, 

 1883, five gave butterflies {Neglecta) on 14th, loth, IGth July, at ten and eleven 

 days' pupation. On 27th November, four were dead, but showed full color, 

 Ic? 3?. The remaining sixteen are alive 14th January, 1884. In former years, 

 chrysalids from Ciinicifuga have given butterflies on 13th, 15th, 17th July ; the 



