GENUS CASTNIA AND SOME ALLIED GROUPS. 163 



species of Lepidoptera the number of branches of the veins in the hind wings is smaller 

 than in the fore wings ; but it will suffice to examine the wings of Hepialm humuli 

 (Pl. XXIX. figs. 27 & 28) to be convinced that the same elements normally exist in 

 the hind as in the fore wings. It will be seen from the lettering of the branches, that 

 the costal, subcostal with its six branches, median with its four branches, and anal veins 

 are present ; and even the discoidal cell is cut up into the three discoidal, subdiscoidal, 

 and premedian divisions. This will enable us to determine satisfactorily the normal 

 condition of the veins and branches when some of the latter are not present in the hind 

 wings. 



In C. cacica (Pl. XXYIII. fig. 6 a) the subcostal vein has only two branches, the 

 second of which is united with the base of c 3* by a very thin curved fold, whilst the 

 veinlet connecting the base of c 3 with that of c 3* extends backwards, and is continued 

 into a discoidal nervure extending nearly to the base of the wing. 



In C.juturna (Pl. XXIX. fig. 1) the same arrangement exists, except that the thin. 

 fold arising a little beyond the base of 6 2 and c 3* is strongly angulated, and a very 

 slight longitudinal fold extends from the angle towards tlie base. 



In C. cochrus (Pl. XXVIII. fig. 11 a) the fold connecting the extreme base of c 2 

 and c 3* is aknost obsolete, but is seen to spring from the discoidal vein at some distance 

 before the origiu of c 3*, thus leaviug the discoidal cell nearly entirely open. 



In C. Hubneri (Pl. XXVIII. fig. 14a)the discoidal is fairly divided into three portions, 

 namely, the discoidal, subdiscoidal, and premedian portions. 



In C. paradoxa the vein or fold connecting the branches b 2 and c 3* is entirely 

 wanting, and the true discoidal cell is wanting ; but its premedian portion is present, 

 in consequence of the branch c? 3* being extended backwards, nearly to the base of the 

 wing. 



C. acrceoides (Pl. XXIX. fig. 14) agrees with C. paradoxa, except that c 2 and c 3 are 

 united together at the base beyond the cell. 



In C. linus (Pl. XXIX. fig. 10) the almost indistinct fold connecting the base of the 

 subcostal vein with the discoidal nervure is pushed still further baek, and thus differs 

 materially in its relation to the branches of the median and subcostal veins. 



In C. mimica the arrangement is the same as in C. linus, except that the transverse 

 fold closing the cell is obsolete, being only indicated by two minute tubercles at the 

 place of its origin (Pl. XXIX. fig. 12). 



In both sexes of C. Marcel Serresi (Pl. XXIX. fig. 3) tlie postcostal vein and its two 

 branches are quite free ; but c 3* extends backwards beyond the base of c 3, forking 

 into two branches, which curve backwards to the base of the wing, thus forming a sepa- 

 rate subdiscoidal cell, the anterior or discoidal part being open, and a very narrow pre- 

 median cell. 



In the female of C. chremes (Pl. XXIX. fig. 5) the same arrangement takes place, but 

 the discoidal cell itself is partially closed by a thin fold ; and in the male the branches 

 c 3 and c 3* arise close together at the extremity of the premedian cell. 



I have thought it advisable to trace these veins and their branches to a normal state 

 in order, if possible, to ascertain their significance, either as sexual, specifie, subgeneric, 



SECONB SERIES. — ZOOLOGT, VOL. I. Z 



