EXTERNAL ANATOMY OF THE MOTH. 49 



the neck. So, too, a median section shows that the neck and prothorax 

 are really one, Plate YIII, Fig. 2. Two pendulous lobes s\ Fig. 1, Plate 

 YIII, project from the upper side of the prothorax, which appear to 

 belong to the scutum, but their real homologies are not yet determined 

 with certainty. They are constricted around their bases, so as to be 

 quite movable. Upon the sides, just behind and below them, lies the 

 first siDiracle, sp^. Below the spiracle are two pieces of elongated shape, 

 running down to the insertion of the coxse; the anterior piece, eps^^ is 

 the episternum, and meets its fellow on the ventral side, forming a me- 

 dian suture between the two coxal articulations, Plate IX, Fig. 1 j the 

 posterior is the homologue of the epimeron of the segments. 



The mesothorax, Plate YIII, Fig. 1, ii, is the largest segment of the 

 body. It is longest on the upper side. The principal dorsal piece is 

 the big scutum, s^^ below the lateral edge of which springs the front 

 wing, tcK Over the base of the wing runs back the large patagium,^^, 

 which is very much developed j its only connection with the body is in 

 front, where it bends over and, as it were, hooks around the front edge 

 of the wing. The scutellum, sm^, is also prominent and extends for some 

 distance over the metathorax. The episternum, eps^, is an oblong 

 piece which runs backward, beginning just underneath the front end of 

 the patagium, and is joined behind to the epimeron, epm^. It is united 

 below with a double ijiece, st, which extends downwards and back- 

 wards to the coxse, ca^. This piece was determined by Dr. Packard, 

 in Attacus, as the sternum, and his nomenclature was followed by Bur- 

 gess in his article on Danais. From our study of the Cotton Moth, it 

 seems doubtful whether this interpretation can be sustained. The piece 

 is double on each side, as can be especially well seen in a ventral view, 

 Plate IX, Fig. 1. The same aspect also shows that the two pieces do 

 not meet in the median line, but are separated by a clear triangular 

 space, beliind which lies a pair (a, a) of pieces which separate the coxae 

 and meet one another in the median ventral line. These are perhaps the 

 real sternal pieces. In any case it is evident that further and extended 

 study is necessary to elucidate the real morphology of these numerous 

 components of the thoracic skeleton. The epimeron, epm^ (Plate YIII, 

 Fig. 1), is quite large and complicated. It consists of a hard Y-shaped 

 piece, between the two legs of which is a large, triangular membranous 

 area. The anterior leg of the Y is much the broader, and joins above 

 to the episternum, and in front to the part marked st, which Packard 

 has held for the sternum. The membranous portion between the legs 

 presents an inconspicuous structure, which is perhaps a spiracle, although 

 this could not be certainly established. The coxal joint, ca?^, tapers 

 rapidly ; examined from the outer surface it appears to consist of two 

 pieces j the anterior piece has been called the coxa, the posterior the 

 trochantine by some authors. 



The metathorax (III) is particularly puzzling on account of its very 

 complex structure, due at least in part to the development of numerous 

 63 CONG 4 



