20 DONALDSON BODINE. 



two different lines of specialization. In one there was a projection 

 of the snrface ventrad, resulting in a form of antenna in which the 

 line of the joints appears to be thrown out from the line of centers 

 of the segments, and is at the dorsal side of the shaft, Fig. 47. 

 When carried very far this projection is emphasized on the mesal 

 line and a distinct ventral ridge is formed. In the other line of 

 specialization, instead of a ventral expansion, there is a sudden pro- 

 jection of a small portion of the segment at the ventro-distal edge 

 forming a flattened expansion whose lateral edges are produced into 

 the more or less slender prolongations which we call pectinations. 



The evolution wrought in the shape of the segments by the earlier 

 and simpler method is a very gradual one. The most generalized 

 forms, e. g. Mleropteryx semipurpurella, see Fig. 1, shows a slight 

 tendency towards the increased development of the ventral surface. 

 In the Microfrenatffi the development is carried further. Here we 

 invariably find a considerable enlargement, but the surface is usually 

 more or less rounded off. Pyrausta oxydalis, Fig. 47, will afford an 

 illustration of what I mean. The stage of the formation of a dis- 

 tinct ventral ridge is best shown in such forms as Cvamhux midahUls, 

 Fig. -1, wliere the ritlge is thin and attains a width equal to twice 

 the diameter of the sliaft pi'oper. This same extreme development 

 is also found among the Sphingida?, Fig. 46. 



The second kind of specialization has undergone a more complex 

 evolution. There can be little doubt that the pectinations arose, 

 perliaps after the first kind of specialization had progressed to a cer- 

 tain extent, as simple extensions of the ventro-distal edges of the 

 segments. The antennae of the females of those species of which 

 the males have highly specialized pectinate antenna? often afford an 

 insight into the probable successive stages through which the forms 

 have attained their high development. The most instructive series 

 for this study can be found among the Satnrniina. All the males 

 of the Citheroniidre and Saturnidse have two pairs of pectinations 

 to a segment for at least a portion of the clavolas, while in only a 

 few genera of the SaturnidiB do the females have them. In most 

 cases the females have non-pectinate antenna? or have only a single 

 pair of pectinations to a segment. Whether they have simply lagged 

 behind the males in the process of specialization, or have degenerated 

 from a common form, it is difficult to say. Mr. Poulton* believes 



•■■ The external Morpholo{?y of the Lepidopterous Pupa ; its Relation to that of 

 the other Stages and to the Origin and History of Metamorphosis. — Part IV, by 

 Edward B. Poulton. Trans. T>inn. Soc, liOndon, second series, vol. v, Zoology, 

 pp. 246- 



