The first half of the paper contains a discussion of the chemical composition of the en velopes of supernovae for the period close to light maximum. The principal conclusions are: The abundance of hydrogen in the envelopes of type I supernovae is low, much lower than that of nitrogen; the abundance of oxygen and carbon is also noticeably lower than that of nitrogen; it seems that there is plenty of helium and metals in these envelopes. The information for type II supernovae is more limited. But

more »

... it is quite certain that the abundance of hydrogen in the envelopes of these stars is much higher than in the envelopes of type I supernovae. In the second half of the paper the problem of supernova remnants is discussed, the circumstellar shells around supernovae (which according to S. van den Bergh and M. Peimbert are ejected from the star before its explosion) are also included. The discussion of this problem permits to confirm again the idea that there is a very close similarity between supernovae and novae. To be more exact there are reasons to suggest that supernovae as well as novae are double star systems, that they are relatively 'old', that they are quite peculiar objects and that they are not the final stage of evolution of 'normal' stars.