If this be admitted, then the elaborate display of their finery by male birds, which Mr. Wallace does admit, may fairly be held to have a value which he does not admit. For if preferential mating is _à priori_ probable, such display may be regarded as the outcome of this mode of selection. At the same time, it may be freely admitted that more observations are required. In a recent paper, "On Sexual Selection in Spiders of the Family _Attidæ_,"[DC] by George W. and Elizabeth G. Peckham, a full, not to say elaborate, description is given of the courtship, as they regard it, of spiders. The "love-dances" and the display of special adornments are described in detail. And the observers, as the result, be it remembered, of long and patient investigation and systematic study, come to the conclusion that female spiders exercise selective choice in their mates. And courtship must be a serious matter for spiders, for if they fail to please, they run a very serious risk of being eaten by the object of their attentions. Some years ago I watched, on the Cape Flats, near Capetown, the courtship of a large spider (I do not know the species). In this case the antics were strange, and, to me, amusing; but they seemed to have no effect on the female spider, who merely watched him. Once or twice she darted forward towards him, but he, not liking, perhaps, the gleam in her eyes, retreated hastily. Eventually she seemed to chase him off the field. We must remember how difficult it is to obtain really satisfactory evidence of mating preferences in animals. In most cases we must watch the animals undisturbed, and very rarely can we have an opportunity of determining whether one particular female selects her mate out of her various suitors. We watch the courtship in this, that, or the other case. In some we see that it is successful; in others that it is unsuccessful. How can we be sure that in the one case it was through fully attaining, in the other through failing to reach, the standard of taste? And yet it is evidence of this sort that Mr. Wallace demands. After noting the rejection by the hen of male birds which had lost their ornamental plumage, he says, "Such cases do not support the idea that males with the tail-feathers a trifle longer, or the colours a trifle brighter, are generally preferred, and that those which are only a little inferior are as generally rejected,--and this is what is absolutely needed to establish the theory of the development of these plumes by means of the choice of the female."[DD] If Mr. Wallace requires direct observational evidence of this kind, I do not suppose he is likely to get any large body of it. But one might fairly ask him what body of direct observational evidence he has of natural selection. The fact is that direct observational evidence is, from the nature of the processes involved, almost impossible to produce in either case. Natural selection is an explanation of organic phenomena reached by a process of