258 ME. E. B. POULTON ON THE EXTERNAL 



DESCRIPTION or THE PLATES. 

 Plate XXVI. 



The Antennae of Lepidopterous Pupse compared with tliose of the corresponding Imagines. 



All the figures are rather smaller than the sizes indicated by the stated amounts of enlargement. The 

 figures were reduced by the lithographer, but the deviation from the originals does not appear to 

 exceed 7 per cent, in any case, and is often much less. 



X 2. The left antenna of a male pupa of Saturnia carpini. The transverse markings are more 



numerou.s than in the figure. 

 X 2. The left antenna of a female pupa of Saturnia carpini. 



X 9. The left antenna of a male imago of Saturnia carpini, as seen from its upper surface, 

 showing the complexity of this sense-organ. Each joint of the antenna is seen to possess two 

 rami upon each side ; these are thickly clothed with sensory hairs (only shown on one side) 

 arranged so as to expose a wide surface to the air. The high degree of development reached by 

 the male antenna corresponds to the size of the pupal antenna within which it is formed (fig. 1). 



Fig. 4. X 9. The left antenna of a female imago of Saturnia carpini, as seen from its upper surface. 

 The organ is very degenerate in size and structure (the latter studied better in figs. 5 and 6), 

 and is out of all ju'oportion to the pupal antenna within which it is formed (fig. 2). The com- 

 parison suggests that tlie degeneration of the sense-organ of the female imago is comparatively 

 recent, so that the pupal organ has not had time to shrink to a corresponding degree. Only a 

 single much dimiuished ramus can be seen on each side of each joint of the antenna, but traces 

 of a second can be made out on some of the joints by the use of a higher power (fig. 5). 



Fig. 5. X 50. Right antenna of the female imago of Saturnia carpini. The sixth and seventh joints 

 above that upon which traces of rami could first be made out in ascending from the basal joint. 

 The joints are represented as seen from below. The longer rami on the left side are directed 

 posteriorly and inferiorly in the natural position of the organ. Thinly scattered sensory hairs 

 are placed upon the rami and the adjacent part of the joints; they are chiefly developed in 

 connexion with the longer rami. In front of (viz., above in the figure) the longer ramus on 

 the lower joint there is a distinct trace of the second ramus in the form of a tubercle bearing 

 a long bristle ; a less distinct tubercle without a bristle is seen on the corresponding part of the 

 upper joint. It is therefore clear that the longer rami correspond to the posterior pair on each 

 joint of the male organ. 



Fig. 6. X 50. Right antenna of another female imago of Saturnia carpini. The joints correspond to 

 those shown in fig. 5, and they are seen from the same point of view. The individual being 

 larger than that from which fig. 5 was taken, the joints are far thicker and larger, although 

 the degeneration has proceeded much farther. Thus the rami are much shorter on both sides, 

 there is no trace of an anterior ramus, and the number of sensory hairs is decidedly smaller, 

 especially on the right side. Tiiis great fluctuation in the degree of degeneration supports the 

 conclusion that the latter is of recent d;itc, a conclusion confirmed by a study of the pupal 

 antenna; (figs. 1 and 2) in relation to those of the imago. 



Fig. 7. X 7. The anterior part of a male pupa of Afflia tau, as seen from the ventral aspect. The 

 enormously broad pupal antenuse occupy a very large area. The character of the surface is 

 only indicated on the left side. 



Fig. 8. X 7. The anterior part of the female pupa of Aglia tau, as seen from the ventral aspect. The 

 pupal antennte are very large, although they do not approach the size reached iu the male. 

 The median ridge is pronounced, and probably corresponds to the part occupied by the 



