An instance illustrating the way in which bees sometimes make provision beforehand of a place to fly to when about to swarm came under our own notice a few years since. A lady who lived about a quarter of a mile from our apiary sent to us to say that a swarm had gone in at a hole over her stable, and to ask us to come and hive them. On our going to do so her gardener told us that he had seen three days previous two or three bees as if reconnoitring; next day several came, and about eleven o'clock on the third day the whole swarm went in and took up their position between the rafters under the flooring. The difficulty was now to get at them. A carpenter was sent for, the boards were taken up, a hive was set over, with a brood comb placed in it to attract them, and by dint of smoke and brushing with a feather, the queen and her retinue were coaxed to ascend into the hive. Some of the bees had already gone out to forage, and there were many flying about that had not settled; so to secure these and make it easy for them, we brought the hive out, and erected a sort of platform on a pair of steps close to the hole, which we stopped. By nighttime all the out-flying bees had joined the swarm and were easily removed. We ought to mention that we recognised this swarm from the appearance of the bees as those from the Carniolan hive left under our care by the Rev. W. C. Cotton (page 45), and as the queen with the swarm was the original, we had to ask that we might be allowed to take the bees back if we provided a swarm of the ordinary English bees, which offer was accepted. Mr. Cotton eventually took this colony to his residence at Frodsham near Chester, and we kept the stock, which of course had a new queen. The bees did not long retain their distinctive features. A swarm of bees, in a natural state, contains from 10,000 to 20,000 insects. "On an average," says Dzierzon, "we may call 20,000 a strong swarm, 12,000 to 15,000 a moderate one, and 6,000 to 8,000 a weak one." Von Berlepsch by a very careful experiment estimated that there were 5,600 unloaded bees in a pound, so that when loaded for swarming there would certainly not be more than 4,000. A good swarm will therefore weigh from three to five pounds. We have known swarms not heavier than two pounds and a half that were in very excellent condition in August as regards store for the winter; though the Baron's experiments showed with remarkable conformity that for a new swarm _six_ pounds was proportionally more profitable than any other weight, larger or smaller. For a fully furnished hive, he states, there seems really no limit but that of space--the more bees the better.