I put a queen, at the moment of her last metamorphosis, into a hive well stored, and sufficiently provided with workers and males; the entrance was contracted so as to prevent her exit, but allowed free passage to the workers. I also made another opening for the queen, and adapted a glass tube to it, communicating with a cubical glass box eight feet high. Hither the queen could at all times come and fly about, enjoying a purer air than was to be found within the hive; but she could not be fecundated; for though the males flew about within the same bounds, the space was too limited to admit of any union between them. By the experiments related in my first letter, copulation takes place high in the air only: therefore, in this apparatus, I found the advantage of retarding fecundation, while the liberty the queen now had, did not render her situation too remote from the natural state. I attended to the experiment fifteen days. Every fine morning, the young captive left her hive; she traversed her glass prison, and flew much about, and with great facility. She laid none during this interval, for she had not united with a male. On the sixteenth day, I set her at liberty: she left the hive, rose aloft in the air, and soon returned with full evidence of impregnation. In two days, she laid, first the eggs of workers, and afterwards as many as the most fertile queens. It thence followed, 1. That captivity did not alter the organs of queens. 2. When fecundation took place within the first sixteen days, she produced both species of eggs. This was an important experiment. It rendered my labours much more simple, by clearly pointing out the method to be pursued: it absolutely precluded the supposed influence of captivity; and left nothing for investigation but the consequences of retarded fecundation. With this view, I repeated the experiment; but, instead of giving the virgin queen liberty on the sixteenth day, I retained her until the twenty-first. She departed, rose high in the air, was fecundated, and returned. Thirty-six hours afterwards, she began to lay: but it was the eggs of males only, and, although very fruitful afterwards, she laid no other kind. I occupied myself the remainder of 1787, and the two subsequent years, with experiments on retarded fecundation, and had constantly the same results. It is undoubted, therefore, that when the copulation of queens is retarded beyond the twentieth day, only an imperfect impregnation is operated: instead of laying the eggs of workers and males equally, they will lay none but those of males.