"As to the stories told of the _Mygale_ catching and killing birds, I am satisfied, both from inquiry and observation, that, at least in Ceylon, they are destitute of truth, and that (unless in the possible case of acute suffering from hunger) this creature shuns all description of food except soft insects and annelides." And yet he immediately adds:--"A lady at Marandan, near Colombo, told me that she had, on one occasion, seen a little house-lizard (_gecko_) seized and devoured by one of these ugly spiders."[159] Does he not, then, credit his informant? Or are lizards included in the category of "soft insects and annelides?" Against this incredulity, resting on no better than negative evidence, one might adduce collateral proof from analogy. There _are_ spiders which feed on vertebrate animals, and there _are_ spiders whose webs catch birds. The large and beautiful _Nephila claripes_ of tropical America weaves strong threads of yellow silk in the paths of the woods, converging to a web quite strong enough to arrest a bird of weak flight. It must have been a species allied to this, but certainly, I think, not the same, of which Dr Walsh speaks in his "Travels in Brazil." "Among the insects is an enormous spider, which I did not observe elsewhere. In passing through an opening between some trees, I felt my head entangled in some obstructions, and on withdrawing it, my straw hat remained behind. When I looked up I saw it suspended in the air, entangled in the meshes of an immense cobweb, which was drawn like a veil of thick gauze across the opening, and was expanded from branch to branch of the opposite trees as large as a sheet ten or twelve feet in diameter. The whole of this space was covered with spiders of the same species but different sizes; some of them, when their legs were expanded, forming a circle of six or seven inches in circumference.[160] They were particularly distinguished by bright spots. The cords composing the web were of a glossy yellow, like the fibres of silkworms, and equally strong."