( 154 ) 



Tribe Sphing'ulicae nov. 



cJ ? . End-segment of antenna short, not prolonged into a thin filiform process. 

 Tongue half the length of the bod}-, or shorter. Spinnlation of abdomen weak, 

 spines of sternites absent, or as weak as scales. No patch of friction-scales on 

 clasper. Rlesotarsus witliont basal comb. Paroiiyohinm witli two lubes at each 

 side, or without lobes, never with one lobe. 



Uab. Oriental Region. 



This small group of genera sliows many affinities with the AmhiiUciimc in the 

 imago state as well as in the larva and pupa, and takes an intermediate position 

 between the tribe Spliiiiqirae and the subfamily Amhidiclnac. Of special interest 

 is the peculiar development of the claw-segment in this tribe. As in the Sphinyicae, 

 the paronychium has in the lower genera two lobes at each side ; while, however, 

 in the SpltiiHi'tcde the next stage in the retrogressive development of the paronychium 

 is rejiresented by a great number of genera with the ventral lobe aborted and tlie 

 subdorsal lobe preserved, no such stage is found in the Spliingulicae, where both 

 lobes disapiicar at the same time, as exemplified by Sphingulns. Another remark- 

 able fact is illustrated by this latter geuus. Among the Sjj/iinc/icne the 'pulvilhi.i 

 d;sa])pears always before the lobes of the paronychium ; there is no s]iecies of 

 Sjihiiigirae which has a pulvillus and is devoid of a paronychial lobe. In Spliingiiliis, 

 however, the paronychial lobes become obliterated, before the pulvillus shows any 

 sign of reduction.* That is to say, we find exactly the ojijiosite line of development 

 in the two tribes. If this holds good for all the Spliinqulleae, as we believe it does, 

 then Dolbiiwpiiis and Uoplioniemn^ which are both without pulvillus and paro- 

 nychial lobes, liave first lost the paronychial lobes and then the pulvillus. The 

 intermediate stage between Synoeclia and Hopliocmma, and between Dolbina. and 

 Jhlb/iiopsis, had pulvilli but no paronychial lobes. Such a stage, corresponding to 

 Sjiliine/iilus, is not known, or, we are inclined to say, not yet known. 



In the diagram inserted opposite, the genera are tabulated according to the 

 relationship. It will be noticed that the whole tribe is divided into one group of 

 genera with the fore- and midtibiae spinose, and another group with the tibiae 

 not spinose. The latter group consists of two branches, of which the two more 

 generalised genera Dolbimi and Keiitoclirysalis are very closely allied, while the 

 specialised genera Sjiliinguhin and Dolbinopsis, which head the two branches, are 

 widely different from one another. Dolhinopms, though clearly a derivation from 

 Dolbina, agrees in several characters with the most specialised genus Ilopliocnema 

 of the branch with sjiinose tibiae, namely in negative characters which are the 

 outcome of the retrogressive development so often referred to among the Sphingimie 

 and Ambulicinae. Both Dolbinopsis and Ilopliocnema are devoid of pulvilli and 

 ]iaronychial lobes, and have the foretibia armed with a thorn (not a spine). 

 This similarity is the result of parallel development, and does not at all indicate 

 close relationship. Further, we find the bristles of the pilifer in a normal state 

 in Tetracliroa and Sgnoecha, while they are reduced in number or developed 

 to hair-scales in Ilopliocnema and all the genera of the branch with uon-si)iuose 

 tibiae. We meet with the same peculiar modification also among the Ambulicinae. 

 This again does not mean closer relationshij) between the genera with scaled pilifer. 



The larva and pnjja, so fiir as known, agree in many respects with those of the 

 tyi)ical Ambulieinne. 



' The s;imc is the case in some Amhiitinime. 



