76 Miscellaneoics. 



various Oolletes ; it lives, therefore, in the larval state at the 

 expense of those Hymenoptera. 



Several reflexions deserve to be recorded here upon this matter : — 



1. The pseudochrysalides, which I found in considerable numbers, 

 were not enclosed in the cells of the Hymenoptera, but lay in their 

 vicinity, in the sand. My rearings give me the reason of this fact, 

 — I have always found (and M. Lichtenstein had also pointed this 

 out) that the second larva of the Caniharis, after exhausting its 

 provision of honey, buried itself in the soil, to become converted 

 there into the pseudochrysalis. Things go on in nature as in my 

 experimental tubes, and I find the pseudochrysalis among the sand, 

 at a greater or less distance from the ceU in which the larva lived 

 as a parasite. This is a distinctive peculiarity which also belongs to 

 the Cerocoma, and constitutes an important differential character 

 with regard to the mode of development of Sitaris and Stenoria. 

 The latter remain to the close of their evolution within the cells of 

 which they have taken possession, and it is in these cells that we 

 find these pseudochrysalides. The larva? of Gantliaris, like those of 

 Cerocoma, are sufiiciently powerfully armed to explain easily how, 

 after having exhausted the provision of honey, they succeed in 

 perforating the very thin wall of the cells of the CoUetes in order to 

 bury themselves in the sand. 



2. I think I may repeat with regard to Gantharis what I said of 

 Cerocoma. I have found the pseudochrysalis of Gantharis in the 

 midst of the cells of CoUetes, but I do not think that these Hyme- 

 noptera are the only ones capable of nourishing the parasitic larvae. 

 The various subterraneous Hymenoptera which provide their larvse 

 with a pasty honey may be indiff'erently the hosts of these parasites, 

 and in proof of this I have the artificial rearings. M. Lichtenstein 

 succeeded in rearing the larvae of Gantharis by feeding them on the 

 honey of Ceratina. I have also succeeded by means of the honey 

 of Megachile and that of Osmia tridentata. 



3. It seems probable, considering the comparatively small size of 

 the ceUs of CoUetes signata, that, in order to arrive at its full deve- 

 lopment, the Gantharis must consume the honey of several cells. It 

 is easy to understand that this may be the case when one knows 

 the voracity with which the larva? of this insect devour honey and 

 the activity that they display. 



In conclusion, I shall record an experiment that I have made for 

 tho purpose of destroying once for all the idea put forward by 

 Nentwich, who asserts that the vesicating power of the Cantharides 

 is only developed after copulation. I have already shown that the 

 cantharidiue has its chief locality in the generative organs ; and I 

 have taken advantage of an opportunity that occurred to me of 

 studying the action of those organs, with perfect certainty that there 

 had been no copulation, since the insect had attained the perfect 

 state under my own eyes. I therefore removed the generative 

 organs from the insect on the 7th June at 11 o'clock in the morning, 

 and applied them immediately upon my forearm after the method 

 which I have already indicated. At 5 o'clock in the evening the 

 apparatus was removed, and a considerable vesicle was soon deve- 

 loped. This experiment can leave no doubt as to the error committed 

 by Neutwich. — Comptes Sendus, June 8, 1885, p. 1472. 



