DYAR: LARVAE OF THE HIGHER BOMBYCES. 
139 
iii on thorax. The abdominal warts are normal, iv and v both very 
small, but distinct. The hairs are spreading, Arctian-like. 
Cosmosoma auge, Linn. 
The warts are arranged as above, but iv and v are of nearly 
normal size. The hairs are of very even length, even the tufted 
plumes which arise from warts iii and iv on joints 5 and 11 do not 
differ in length. Warts small, neatly defined, and contrasting in 
color with the pale body. The single subdorsal wart on the thorax 
is no longer than the others. 
Stage I is like Scepsis and illustrates the same facts. 
Family Pericopidae. 
The larvae of this family have all the essential wart structure of 
the Arctiidae. I have elsewhere discussed the only larva known 
to me and given reasons for supposing the family to represent a 
group of specialized Dioptids. (See Journ. N. Y. ent. soc., vol. 4, 
p. 68.) 
Family Eupterotidae. 
Following the suggestion of Mr. W. Schaus, I include the genus 
Ajnatelodes in this family; also the other hairy Notodontians, 
Melalopha, Datana, and Phalera 1 . Recently I discussed the evidence 
for placing. Apatelodes in this group (Can. ent., vol. 27, p. 159) 
and showed that the characters of the mature larva did not contra¬ 
dict it. In a paper in Psyche (vol. 7, p. 316-317) I give my obser¬ 
vations on the first stage of Apatelodes, showing distinctly the wart 
characters of the Eupterotidae. I have already discussed several 
species of Thaumetopoea Hbn. (= Cnethocampa) (Journ. N. Y. 
ent. soc. vol. 3, p. 22-23.) 
I have described the formation of the three upper thoracic warts 
to be from primitive tubercles ia, ib-fiia, and iib( Trans. N. Y. acad. 
sci., vol. 14, p. 57) ; but I do not find positive evidence of the exact 
way in which they were derived. Whether it be in this way or as 
in the Arctians (as is indicated by the arrangement of setae in stage 
I of the Notodontian Schizura unicornis ), the lower wart of the 
i It appears that Melalopha and Apatelodes might be properly classed as Eupterotidae 
on imaginal characters since they have no tongue. Datana has a rudimentary one, but 
I have given more weight to the larval characters than to the persistence of this ves¬ 
tigial structure. 
