I am inclined to think, that this effluvium is not altogether an imaginary phenomenon. It seems to me to exclude, at least, the hypothesis of the persistence of the retinal image; for the false image does not last so long as the effluvium, under the conditions mentioned by me. There is yet another explanation. This is that the eye automatically prolongs the clear impression of the fingers on the dark background separating them. This would be analogous to the expansion by irradiation of clear images upon a dark background. Other reasons, however, make me discard this hypothesis. In the first place, why do some people see the supposed false image vividly coloured and not white? Secondly, if the phenomenon is of retinal origin, why—instead of being thinner, as is the case—does not the image reproduce the form of the finger? Why is it a blue-grey colour and not black, as should be the complimentary image of a finger which appears to be white? Why is not the phenomenon produced with certain objects coloured in white? In vain might we experiment with them as with the hands; they would never leave effluvium between them. There is an exception, however: if we hold cotton or wood in the hands, we will often perceive this appearance of effluvium. It is not obtained, as far as I have been able to judge, with metal objects. From this, it may be inferred, though I do not affirm it, as my experiments are not sufficiently conclusive—that wood and cotton conduct the effluvium as well as flesh. This seems to me very probable with cotton; by holding a crumpled handkerchief in the hand, and presenting it to the background as I have recommended doing with the fingers, we will notice a slight mist round the cotton, which seems to soften off the outlines. Finally, another more serious reason for considering these effluvia as probably objective, is the frequent absence of parallelism between the effluvia of corresponding fingers. I have often observed distinct divergencies, and it sometimes struck me as though the will might be able to influence the direction of the effluvia to a certain extent. It often happens that all the experimenters see the effluvia under the same aspect. The phenomenon can show great variability in appearance, the middle finger of one hand, for example, becoming connected with two, three, or four fingers of the opposite one. As the aspect of this effluvium usually appears the same to the observers, there is room to presume that its existence and direction are not illusory phenomena. In the contrary case, we would have to suppose collective hallucination, or a most improbable transmission of impression, which my personal observations do not dispose me to admit.