94 



THE ENTOMOLOGIST. 



with the later stages, I believe that a series of superfamilies 

 can be satisfactorily defined. This is as far as classification 

 from the arrangement of the tubercles can be carried. . . . 

 Superfamily Cossina (Micro-Lepidoptera) : Tubercles with single 

 seta, normal, fourth and fifth approximate or consolidated, the 

 rest remote Includes the families Adelidas, Psychidse, Cos- 

 sidas, Pyralidffi, Tortricidae, Sesiidte, Tineidse, Orneodidae, and 

 Lacosomidae " 



The ovum of C. cossns has been described by Mr. J. W. Tutt, 

 F.E.S, (Ent. Rec. xv. 3B3), and to Mr. A. Bacot, F.E.S., I am 

 indebted for specimens of newly-hatched C. cossm larvae ; these 

 are quite large animals, in length 3"5 mm. I believe the species 

 has not hitherto been figured in this larval stage, and some 

 remarks on the homology of the thoracic tubercle setae with 

 abdominal tubercle setae are suggested by examination of these 

 newly-hatched larvae of C. cossus. 



Looking further afield than Cossidae, one notes that a fun- 

 damental arrangement of abdominal setae, of newly-hatched 

 larva3 especially, largely prevails throughout the Lepidoptera. 

 Approximation and loss of certain tubercles is admitted to be 

 specialization. In Hepialidae, newly-hatched larvae have this 

 fundamental arrangement also on the thoracic segments — i. e. 

 anterior first, posterior second, supraspiracular third, subspira- 

 culars fourth and fifth, and basal — correspond in number and 

 position with the abdominal tubercle setae. It is here necessary 

 to note that I refer only to the meso- and post-thoracic segments, 

 the prothorax everywhere has a special arrangement of the setae. 

 In Hepialidae* iii is applied to two setae — in fact, iiia, b, the 

 spiracle, and iv, v, tubercles, are curiously associated on one 

 common area in the newly-hatched stage. 



As indicated by Hepialidae, no doubt the primitive arrange- 

 ment of the tubercle setae was alike on both thoracic and abdo- 

 minal segments. Movement (noticably approximation) of setal 

 tubercles has taken place to a greater extent on the thoracic than 

 on abdominal segments. Elimination of the spiracles — outward 

 indication of internal modification in connection with wing deve- 

 lopment — has not affected the abdominals, but is probably the 

 chief cause of the altered positions of the tubercles on thoracic 

 segments. 



In lateral descending order the abdominal tubercle setae of 

 newly-hatched C. cossus (figs. 1 and 3) are i and ii normal in 

 position single seta each, iii single seta.t I cannot trace iii Z^, 



- Figure, Trans. Entom. Soc. Lond. 1900. 



I Mr. Bacot believes he can detect iii b on C. cossus. \ Not so ! Mr. 

 Quail uses iii b as denoting the second supraspiracular setae on Hepialidae 

 larvae. The tubercle I remarked on to Mr. Quail as being present on larvae of 

 C. cossus is a minute free spiracular point of very general, if not universal, 

 occurrence on the abdominal segments of lepidopterous larvae, and probably 

 has no relation at all to iii b of Hepialidae. — A. Bacot.] 



