It was discovered by Dzierzon, in 1845, that the drones hatch from unimpregnated eggs. This strange phenomenon, seemingly so incredible, is as has been shown in speaking of the queen, easily proved and beyond question. These eggs may come from an unimpregnated queen, a fertile worker--which will soon be further described--or from an impregnated queen, which may voluntarily prevent impregnation. Such eggs may be placed in the larger horizontal cells (Fig. 28, _a_), in manner already described. As stated by Bevan, the drone feeds six and a half days as a larva, before the cell is capped. The capping of the drone-cells is very convex, and projects beyond the plane of the same in worker-cells, so that the drone brood is easily distinguished from worker, and from the darker color--the wax being thicker and less pure--the capping of both drone and worker brood-cells enable us easily to distinguish them from honey-cells. In twenty-four days from the laying of the egg, the drones come forth from the cells. Of course variation of temperature, and other conditions, as variable amount of diet, may slightly retard or advance the development of any brood, in the different stages. The drones--in fact all bees--when they first emerge from the cells, are gray, soft, and appear generally unsophisticated. Just what the longevity of the male bee is, I am unable to state. It is probable, judging from analogy, that they live till accident, the worker bees, or the performance of their natural function causes their death. The worker-bees are liable to kill off the drones, which they do by constantly biting and worrying them. They may also destroy the drone-brood. It is not very rare to see workers carrying out immature drones even in mid-summer. At the same time, too, they may destroy inchoate queens. Such action is prompted by a sudden check in the yield of honey, and with the drones is most common at the close of the season. The bees seem very cautious and far-sighted. If the signs of the times presage a famine, they stay all proceedings looking to the increase of colonies. On the other hand, unlimited honey, rapid increase of brood, crowded quarters--whatever the age of the queen--is sure to bring many of the male bees. While any circumstances that indicate a future need of drones will prevent their destruction even in late autumn. The function of the drones is solely to impregnate the queen, though when present they may add animal heat. That their nutrition is active, is suggested by the fact, that upon dissection, we always find their capacious stomachs filled with honey.