ATTACID/E. 287 



of i o, II absent. Hindwings without frenulum, la absent, 5 and 6 approximated 

 from upper angle, 7 from | of upper margin, 8 out of cell at base rapidly diverging, a 

 praecostal spur faintly indicated. Ovum : Flattened-spherical, smooth. Larva: Stout, 

 with tubercles bearing tufts of scanty hairs. Pupa : In a dense firm oval cocoon. 



These characters would, in many cases, fail entirely if applied 

 to any group greater than our family Attacidae. Packard characterises 

 four families in the superfamily — Ceratocampidae, Hemileucidae, 

 Agliidae and Saturniidae (= Attacidae). These we accept with the 

 addition of Automeridae. The Attacidae, he subdivides (Proc. Amer. 

 Acad. Arts Sciences, 1893, P- 5&) into two subfamilies as follows : 



1. Six tubercles on the 8th abdominal segment, the tubercles in general 

 over the body, all of the same size. Generalised forms. Subfam. Saturniin/e. 



2. Five tubercles on the 8th abdominal segment, the median one double ; 

 the tubercles in general more or less differentiated or specialised in size and colour. 

 Specialised forms. Subfam. AttaciN/t:. 



Although based on the characters presented by the number 

 and arrangement of the larval tubercles, the same subdivision can 

 be made on other characters, and it appears to us to be a very- 

 natural one, but Packard's remark, "generalised forms," is quite inex- 

 plicable to us. The family, as here restricted, contains the large 

 silkworm moths, whose silk is used for commercial pui poses, e.g., 

 Antheraea mylitta and its allies (India), A. pernyi (China), A. 

 yamamai (Japan), Callosamia promethea, Samia cecnpia (North 

 America), Philosamia lunula (East Indies), and their allies. The 

 family contains some of the largest moths in the world ; the largest, 

 Attacus atlas, sometimes reaching a foot in expanse. 



Dealing with the origin of the family, Packard states 

 (Proc. Am. Acad. Arts and Set., 1893, p. 55): "The larval 

 characters, especially those features that are congenital, tend 

 to show that the family has originated from some spiny group, 

 and most probably, when we take into account the transfor- 

 mations of Aglia tau, from the Ceratocampidae ( ' Citheroniidae ), 

 although none of the latter spin a cocoon. During the evolution 

 of the group they underwent a change in shape, from a rather long 

 and slender form to a thick heavy body, with a thin integument, 

 the result perhaps of an unusually stationary mode of life. The 

 imagines also underwent a process of degeneration, as seen in the 

 atrophy, total or partial, of the maxillae, and in the loss of veins, 

 in their large but weak wings*; though the loss of strength of flight 

 is somewhat compensated by the remarkable development of the 

 olfactory organs or antennae. The family appears to be a closed 

 type, none of the higher or more specialised Bombyces having 

 descended from it, the type representing a side branch of the 

 Bombycine tree which, late in geological history, grew apart, and 

 reached a marked degree of modification, resulting in the possession 

 of adaptive characters which were not transmitted to later forms. 

 It seems probable that the type was a Miocene-Tertiary one, which 

 has lingered on in eastern America (north and south), and in 

 eastern Asia, as well as in Africa, while it has become nearly 

 extinct on the Pacific shores of North and South America, and in 

 Europe." Packard further notes {Bombycine Moths, etc., p. 45) that 

 "the Saturniidae {Attacidae) is a tropical group, only a single 



* Bacot notes: "The movement of the wings of Attacus atlas is slow and 

 bat-like, but hardly weak, and the wing-membraue must be very strong, as it stands 

 well a considerable amount of knocking about in a confined space." 



