the Larva of Collyris emarginatus. 85 
short decurved spines that have already been mentioned.* 
The third segment is composed of the lateral sclerites, 
and a median dorsal sclerite, which runs backward and 
ends between the base of the crotchets (Plate ITI, fig. 9. d) ; 
the lateral sclerites meet each other in the mid-ventral 
line. The fourth segment is open ventrally, the lateral 
sclerites are now pillars bearing the crotchets and each has 
a small hook on the outer aspect, the dorsal sclerites 
are represented perhaps by a pair of oval setigerous plates 
(Plate III, fig. 9. sp.) covering the base of the crotchets. 
Each crotchet consists of three strong curved hooks, the 
second of which has on the ventral aspect an inwardly 
projecting flange (Plate III, fig. 10.7); they are articu- 
lated to the lateral sclerites by a transverse joint but, 
so far as I know, are not movable independently of the 
chitinous tube. This is all that can be made out from an 
examination of dried specimens and I am unable to afford 
any information as to the exact relations of these parts to 
the other internal organs of the beetle. But there can be 
no reasonable doubt that the segmented chitinous tube is 
composed of retracted terminal segments, the last one of 
which bears appendages in the form of crotchets, and it is 
these appendages only which can be regarded as the 
morphological equivalents of the female gonapophyses of 
the Terebrant Hymenoptera. The modus operandi of the 
genital armature of Collyris is obscure, but I have little 
doubt of its efficiency as an instrument for boring through 
wood of no greater hardness than young coffee twigs. 
Strictly homologous organs occur in other Cicindelidee and 
doubtless in every case they function as boring tools. So 
far as is known—though observations on the subject are 
woefully inadequate—the Cicindelidee deposit their eggs im 
substances, and not on surfaces, and it does not require a 
great stretch of imagination to suppose that the arboreal 
Collyris only departs from the habits of its allies so far as 
to deposit her eggs inside the twigs of trees and shrubs. 
It is of interest to note that the pair of decurved ventral 
spines are only well-developed in the arboreal species,+ 
* These spines have been described elsewhere as attached to the 
last visible sternite, but this is manifestly incorrect. 
+ Wallace states that Therates labiata in Amboina is arboreal and 
in this species the ventral spines are well developed ; in other species 
that I have examined these spines are minute or absent, and Canon 
Fowler informs me that occasionally they are modified to form comb- 
