The quarries form part of an extensive plateau which is so arid as to be nearly deserted. In these conditions, the Osmia, at all times faithful to her birth-place, has little or no need to emigrate from her heap of stones and leave the shell for another dwelling which she would have to go and seek at a distance. Since there are heaps of stone there, she probably has no other dwelling than the Snail-shell. Nothing tells us that the present-day generations are not descended in the direct line from the generations contemporary with the quarryman who lost his as or his obol at this spot. All the circumstances seem to point to it: the Osmia of the quarries is an inveterate user of Snail-shells; so far as heredity is concerned, she knows nothing whatever of reeds. Well, we must place her in the presence of these new lodgings. I collect during the winter about two dozen well-stocked Snail-shells and instal them in a quiet corner of my study, as I did at the time of my enquiries into the distribution of the sexes. The little hive with its front pierced with forty holes has bits of reed fitted to it. At the foot of the five rows of cylinders I place the inhabited shells and with these I mix a few small stones, the better to imitate the natural conditions. I add an assortment of empty Snail-shells, after carefully cleaning the interior so as to make the Osmia's stay more pleasant. When the time comes for nest-building, the stay-at-home insect will have, close beside the house of its birth, a choice of two habitations: the cylinder, a novelty unknown to its race; and the spiral staircase, the ancient ancestral home. The nests were finished at the end of May and the Osmiae began to answer my list of questions. Some, the great majority, settled exclusively in the reeds; the others remained faithful to the Snail-shell or else entrusted their eggs partly to the spirals and partly to the cylinders. With the first, who were the pioneers of cylindrical architecture, there was no hesitation that I could perceive: after exploring the stump of reed for a time and recognizing it as serviceable, the insect instals itself there and, an expert from the first touch, without apprenticeship, without groping, without any tendencies bequeathed by the long practice of its predecessors, builds its straight row of cells on a very different plan from that demanded by the spiral cavity of the shell which increases in size as it goes on.