Before I finish with _C. imperialis_, I must mention a peculiar fact. During the first stage, and, I think, also during the second, several larvæ disappeared without leaving any traces. I also saw two smaller larvæ held tight by the hind claspers of two larger ones. The larvæ thus held and pressed were perfectly dead when I observed them, and I removed them. My impression then was that these larvae were carnivorous, not from this last fact alone, as I had previously observed it with larvæ of Catocalæ when they are too crowded, but from the fact that some had disappeared entirely from the glass under which they were confined. I began to reduce their numbers, and put six only under each glass, so as to be able to watch them better. Whether I had made a mistake or not previously to this I do not exactly know; but from this moment the larvae behaved in a most exemplary manner, especially when they became larger. They crawled over each other's backs without the least sign of spite or animosity, even when they were in sleep, in which case larvæ are generally very sensitive and irritable, all were of a most pacific nature. It is, therefore, with the greatest pleasure that, for want of sufficient evidence, I withdraw this serious charge of cannibalism which I first intended to bring against them. From what has been said respecting the rearing of exotic silk-producing bombyces, especially tropical species, it must have been observed that several difficulties, standing in the way of success, have to be overcome. The moths of North American species emerge regularly enough during the months of May, June, or July, but Indian and other tropical species may emerge at any time of the year, if the weather is mild, as has been the case during this unusually mild winter of 1881-1882. From the end of December to the present time (March 14, 1882) moths of four species of Indian silk-producers, especially _Antheræa roylei_ and _Actias selene_, have constantly emerged, but only one or two at a time. These moths emerged from cocoons received in December and January last. It is only when these tropical species shall have been already reared in Europe that the emergence of the moths will be regular; then they will be single-brooded in Northern or Central Europe, and some will very likely become double-brooded in Southern Europe. But when just imported the moths of these tropical species will always be uncertain and irregular in their emergence; hence the importance of having a sufficient number of cocoons so as to meet this difficulty, i.e., the loss of the moths that emerge prematurely or irregularly.