The stone support is the most frequent, though not the only one. I have found nests, but sparsely inhabited it is true, on the trunks of trees, in the seams of the rough bark of oaks. Among those whose support was a living plant, I will mention two that stand out above all the others. The first was built in the grooves of a Peruvian torch-thistle as thick as my leg; the second rested on a stalk of the opuntia, the Indian fig. Had the fierce armour of these two stout cactuses attracted the attention of the insect, which looked upon their tufts of spikes as furnishing a system of defence for its nest? Perhaps so. In any case, the attempt was not imitated; I never saw another installation of the kind. There is one definite conclusion to be drawn from my two discoveries. Despite the oddity of their structure, which is unparalleled in the local flora, the two American importations did not compel the insect to go through an apprenticeship of groping and hesitation. The one which found itself in the presence of those novel growths and which was perhaps the first of its race to do so took possession of their grooves and stalks just as it would have done of a familiar site. From the start, the fleshy plants from the New World suited it quite as well as the trunk of a native tree. The Mason-bee of the Pebbles (Chalicodoma parietina) has none of this elasticity in the choice of a site. In her case, the smooth stone of the parched uplands is the almost invariable foundation of her structures. Elsewhere, under a less clement sky, she prefers the support of a wall, which protects the nest against the prolonged snows. Lastly, the Mason-bee of the Shrubs (C. rufescens, Pérez) fixes her ball of clay to a twig of any ligneous plant, from the thyme, the rock-rose and the heath to the oak, the elm and the pine. The list of the sites that suit her would almost form a complete catalogue of the ligneous flora. The variety of places where the insect installs itself, so eloquent of the part played by discernment in their selection, becomes still more remarkable when accompanied by a corresponding variety in the architecture of the cells. This is more particularly the case with the Three-horned Osmia, [54] who, as she uses clayey materials very easily affected by the rain, requires, like the Pelopæus, a dry shelter for her cells, a shelter which she finds ready-made and uses just as it is, after a few touches by way of sweeping and cleansing. The homes which I see her adopt are especially the shells of Snails that have died under the stone-heaps and in the low, unmortared walls which support the cultivated earth of the hills in shelves or terraces. The use of Snail-shells is accompanied by the no less active use of the old cells of both the Mason-bee of the Sheds and of certain Anthophoræ (A. pilipes, A. parietina and A. personata.) [55]