But there is another situation where we can judge of the influence of the royal food administered to worms in common cells. I have spoken at great length in my letter on the existence of fertile workers. You cannot forget, Sir, that the expansion of their sexual organs is owing to the reception of some particles of royal jelly, while in the vermicular form. For want of new observations, I must refer you to what is previously said on the subject. _PREGNY, 4 September 1791._ LETTER IX. _ON THE FORMATION OF SWARMS._ I can add but a few facts to the information M. de Reaumur has communicated relative to swarms. A young queen, according to this celebrated naturalist, is always or almost always at the head of a swarm; but he does not assert the fact positively, and had some doubts on the subject. "Is it certain," says he "as we have hitherto supposed, in coincidence with all who have treated of bees, that the new colony is always conducted by a young mother? May not the old mother be disgusted with her habitation? or may she not be influenced by some particular circumstances to abandon all her possessions to the young female? I wish it had been in my power to solve this question otherwise than by mere probabilities, and that some misfortune had not befallen all the bees whose queen I had marked red on the thorax." These expressions seem to indicate, that M. de Reaumur suspected that the old queens sometimes conducted the young swarms. By the following details, you will observe, that his suspicions are fully justified. In the course of spring and summer, the same hive may throw several swarms. The old queen is always at the head of the first colony; the others are conducted by young queens. Such is the fact which I shall now prove; and the peculiarities attending it shall be related. But previous to entering on this subject, I should repeat what has already been frequently observed, that the _leaf_ or flat hives are indispensible in studying the industry and instinct of bees. When they are left at liberty to conduct several rows of parallel combs, we can no longer observe what is continually passing between them, or they must be dislodged by water or smoke, for examining what has been constructed; a violent proceeding, which has a material influence on their instinct, and consequently exposes an observer to the risk of supposing simple accidents permanent laws. I now proceed to experiments proving that an old queen always conducts the first swarm.