VIII 



the Hesperidae are nearest allied, the answer cannot be doubtful; the above 

 named frenulum, as has been said, does indeed occur in this family, but only 

 very rarely, and evidently as a mere relic ; and their antennae are of the same 

 stiff thread-like nature as those of the Rhopalocera, though the club-form is 

 somewhat modified; with those of the Heterocera they agree in no wise. All 

 this is still further corroborated when with regard to affinity, we do not, as is 

 usual in systematism, merely pay attention to the characteristics of the imago, 

 but also biologically take into account the early stages. The Rhopalocera have 

 undoubtedly become differentiated from the Heterocera and specially from those of 

 whose caterpillars the spinning-power was strongly developed. This we infer from 

 the circumstance that this power is not merely present in the first named group 

 as a mere relic, but can as such be clearly seen in different degrees of gradual 

 decrease by any one who is convinced that for the right understanding of 

 structural changes it is necessary, as I have already pointed out so emphatically 

 in my considerations on the phenomenon of colour-evolution, always to bear in 

 mind their evolutional character. It is not accidental that in some families of the 

 Rhopalocera, the spinning power of the larvae when pupating is still manifest 

 only in the attaching of the anal extremity of the pupa by a button of silk, 

 while in a great number of families the pupae are attached, besides at the 

 anal extremity, also by a so-called silken girdle-thread, which is evidently 

 purposely spun for this. Undeniably the reason of this is an evolutional decay 

 in the use of the spinning power, as I have already pointed out on page 72 

 of my monography on the Javanese Pieridae ') which may be ascribed to the 

 circumstance that where the animal had slowly, as if through involuntary practice, 

 acquired the power to pass the pupal state with the head extremity turned 

 downwards, it was no longer necessary for the pupa to be kept in an upright 

 posture by means of the girdle-thread, so that consequently the girdle-thread had 

 become superfluous. This simplification, however, has only been accomplished 

 in a small number of families, although in the others also there is often a 

 manifest tendency in the same diretion. In the latter group, the succincti of 

 BoisDUVAL, the spinning power results therefore in so much more as is required 

 for the application of a girdle-thread, which is no longer necessary in the 

 suspensi. Herewith also agrees the fact that in those cases where in the old 

 forms of Rhopalocera the evolutionary change is as yet much less advanced, 

 and the spinning-power is consequently much more active and agrees more in 

 this respect with those Heterocera from which the Rhopalocera have differentiated, 



") Through a mistake in the translation which was overlooked in the correction of the proofs, 

 the words suspensae and succinetae in lines 2i and 22 were unfortunately transposed. 



