He is even able to leave the family of the Compositae and to go gleaning more or less everywhere. Here is the result of my botanizings at the expense of his nests. We must distinguish between two genera in the Shrike's rough classification: the cottony plants and the smooth plants. Among the first, my notes mention the following: Convolvulus cantabrica, or flax-leaved bindweed; Lotus symmetricus, or bird's-foot trefoil; Teucrium polium, or poly; and the flowery heads of the Phragmites communis, or common reed. Among the second are these: Medicago lupulina, or nonesuch; Trifolium repens, or white clover; Lathyrus pratensis, or meadow lathyrus; Capsella bursa pastoris, or shepherd's purse; Vicia peregrina, or broad-podded vetch; Convolvulus arvensis, or small bindweed; Pterotheca nemausensis, a sort of hawkweed; and Poa pratensis, or smooth-stalked meadow-grass. When it is downy, the plant forms almost the whole nest, as is the case with the flax-leaved bindweed; when smooth, it forms only the framework, destined to support a crumbling mass of micropus, as is the case with the small bindweed. When making this collection, which I am far from giving as the birds' complete herbarium, I was struck by a wholly unexpected detail: of the various plants, I found only the heads still in bud; moreover, all the sprigs, though dry, possessed the green colouring of the growing plant, a sign of swift desiccation in the sun. Save in a few cases, therefore, the Shrike does not collect the dead and withered remains: it is from the growing plants that he reaps his harvest, mowing them down with his beak and leaving the sheaves to dry in the sun before using them. I caught him one day hopping about and pecking at the twigs of a Biscayan bindweed. He was getting in his hay, strewing the ground with it. The evidence of the Shrike, confirmed by that of all the other workers--weavers, basket-makers or woodcutters--whom we may care to call as witnesses, shows us what a large part must be assigned to discernment in the bird's choice of materials for its nest. Is the insect as highly gifted? When it works with vegetable matter, is it exclusive in its tastes? Does it know only one definite plant, its special province? Or has it, for employment in its manufactures, a varied flora, in which its discernment exercises a free choice? For answers to these questions we may look, above all, to the Leaf-cutting Bees, the Megachiles. Reaumur has told the story of their industry in detail; and I refer the reader who wishes for further particulars to the master's Memoirs.