ATTACHE. 289 



absorbed, the primitive cell membrane dies out, and there remains 

 only the chitinised cuticle of the scale, with its root fast in a pouch in 

 the membrane. The striking analogies between the course of develop- 

 ment of scales and the varying conditions of the scales on the fully 

 developed wing cannot escape attention. It is to be noted that 

 Semper's description of the formation of the insertion cups differs 

 from that given by Landois. 



Bodine notes (Antennae of Lepidoptera, p. 43) that the family 

 Saturniidae (Attacidae) is interesting in the series of forms the antennae 

 present. The genera Coloradia, Auto metis, Calosaturnia, Tropaea, 

 Telea, Callosamia, Philosamia and Samia present a regular and 

 progressive series. The following table will serve to separate the 

 antennae of the members of this family * : 



A. Antennse of female with single pair of pectinations to a segment. 



B. Antennae of male with distal pair of pectinations shorter than the proximal. 

 C. Distal pair not more than half the length of the 



proximal . . . . . . . . . . Coloradia. 



CC. Distal pair but little shorter .. .. Automeris. 



BB. Antennas of male with distal and proximal pairs of 



pectinations subequal . . . . . . . . Calosaturnia. 



AA. Antennas of both sexes with two pairs of pectinations to a segment. 



B. Proximal and distal pairs subequal in male, distal pair shorter in the 

 female. 

 C. Distal pair of female very short, without hairs of 



the second type . . . . . . . . Telea. 



CC. Distal pair of moderate length with hairs of 



second type . . . . . . . . . . Iropaea. 



BB. Proximal and distal pairs subequal in both sexes. 



C. Distal pair of pectinations of female shorter than, or only equal to, 

 the proximal on the proximal segments. 

 D. Distal pair wanting in a few distal seg- 

 ments . . . . . . . . . . Callosamia. 



DD. Distal pair present, at least in rudiments, 



to the distal end . . . . . . . Samia. 



CC. Distal pair of pectinations of female longer than 



the proximal on the proximal segments . . Philosamia. 



There is a gradual progression in complexity of development 

 from Coloradia on the one hand, to Samia or Philosamia on the other. 

 The females of Coloradia, Automeris and Calosaturnia, have a single 

 pair of pectinations to a segment, while those of the other genera have 

 two pairs. The males of the first two genera have the distal pair 

 shorter, whilst those of the others have the two pairs subequal. There is 

 also a gradual increase in both sexes from one end of the series to the 

 other in the number and position of the cones. In Samia there is not 

 only an abundance on the shaft, but many are on the pectinations of 

 the distal portion of the clavola. There is some doubt as to whether 

 Samia or Philosamia should be considered the higher form. The 

 male Samia is more highly developed than the male Philosamia, but, 

 on the other hand, the female Samia is less developed than the female 

 Philosamia. I believe that now Samia is the higher form, and that it 

 has outstripped Philosamia in specialisation in comparatively recent 

 times. If the females lag behind the males, as seems most probable, 

 the condition we find in the two genera would indicate that the male 

 Philosamia has been long enough fixed to allow the female to 



* Reference should be made here to Chapman's note on the antennae of African 

 forms [antea, p. 270). 



R 



