In order to ascertain by what means the bee found its way to the hive, whether by seeing it through its compound eyes, or otherwise, Réaumur performed an interesting experiment similar in character to the one before mentioned. He covered with a red varnish, which was quite opaque, the compound eyes of a number of bees taken from the same hive. He then shut them up in a box with several other bees from the same hive which he left untouched. The box was only a few paces distant from the hive from which the bees were taken. He then opened the lid of the box, and those which had not been blinded instantly flew out of it, and entered their habitation. Those whose eyes had been varnished appeared not to care to leave the box at all, and seemed very unwilling to make any attempt to fly; some of them, indeed, flew about from one side to another, but did not go far. Réaumur then threw several of them up into the air, and they immediately began to soar higher and higher, until at length they went out of sight altogether! Réaumur compares the poor insect's manoeuvres to those of a crow, whose head and eyes mischievous boys have covered with a paper bonnet: the bird flies upward until its strength is exhausted, when it drops again to the earth. Not only did those bees which he threw up into the air thus soar until they were lost to view, but all the most active of those which were left in the box did so likewise, and Réaumur saw them no more. Not one could find its way to the hive. From this experiment, and from the preceding one, it is evident that both the compound and the simple eyes are necessary to enable the insect to see perfectly; for when either was varnished over, the bees could not find their way to their home again. Réaumur imagines that the cause of the wheeling flight of bees, sometimes observed, now in this direction, now in the opposite, may arise from their eyes becoming perhaps in part obscured by the pollen, or yellow powder of the flowers into which they plunge, the bee thereby becoming partly blind-folded. These experiments are so interesting that they deserve repeating, and it might be tried whether the result would be different if only one eye were blinded. Other variations of the experiments will also suggest themselves.