Chalicodoma sicula prefers a sociable life to a solitary one, and hundreds—nay, several thousands—will establish themselves on the under surface of the tiles on a hovel, or the edge of a roof. It is not a real society with common interests, dear to all, but merely a gathering where each works for herself and is not concerned for the rest—a throng recalling the swarm of a hive only by their number and industry. They use the same mortar as Chalicodoma muraria, equally resistant and waterproof, but finer and without pebbles. First the old nests are utilised. Every free cell is repaired, stored, and shut up. But the old ones are far from sufficing to the population, which increases rapidly year by year, and on the surface of the nest, where the cells are hidden below the old general mortar covering, new ones are built as required. They are placed more or less horizontally, one beside another, with no kind of order. Every constructor builds as the fancy takes her, where and as she wills; only she must not interfere with her neighbour’s work, or rough treatment will soon call her to order. The cells accumulate in chance fashion in this workyard, where there is no general plan whatever. Their form is that of a thimble divided down the axis, and their enclosure is completed either by adjacent cells, or the surface of the old nest. Outside they are rough, and look like layers of knotted cords corresponding to the layers of mortar. Inside the walls are level but not smooth; a cocoon will replace the absent polish. As soon as a cell is built it is stored and walled up, as we have seen with Chalicodoma muraria. This work goes on through the whole of May. At length all the eggs are laid, and the bees, without any distinction as to what does or does not belong to them, all set to work on a common shelter of the colony—a thick bed of mortar, filling up spaces and covering all the cells. In the end the nests look like a large mass of dry mud—very irregular, arched, thickest in the middle, the primitive kernel of the establishment, thinnest at the edges, where there are fewest cells, and very variable in extent, according to the number of workers, and consequently to the time when the nest was begun. Some are not much larger than one’s hand, while others will occupy the greater part of the edge of a roof, and be measured by square yards. If Chalicodoma sicula works alone, as she often does, on the shutter of an unused window or on a stone or a branch, she behaves in just the same way. For instance, if the nest is on a bough, she begins by solidly fixing the basis of her cell on the slender twig. Then the building rises into a little vertical tower. This cell being stored and ceiled, another follows, supported both by the bough and the first cell, until six to ten cells are grouped one beside the other, and finally a general cover of mortar encloses them all together with the bough, which gives them a firm foundation. XXI EXPERIMENTS