The hypothesis of a psychic automatism of a mediumistic type, as a concomitant phenomenon, at least, in experiments of the "new zoopsychology," offers us a point of support for a possible interpretation of the strange uncertainty and irregularity of the successes and failures of different observers and different animals.

With Krall two of his horses gave magnificent results; two others negative results. In the same way, with the same dogs some experimenters obtain wonders, others obtain nothing.... We may therefore assume that in order to obtain favourable results there must be a proper accord or reciprocal psychic concordance between the animal and the person making the experiment, precisely as happens with mediumistic phenomena.

Moreover, this hypothesis in the same way helps us to an interpretation of the fact that the same animal, with the same investigator, gives good results in some matters, poor or no result in others. Taking, however, due account of the central mathematical phenomena, on which, as it seems to me, the whole edifice is superposed, there remains a great variety of marked psychical idiosyncrasies in the various cases. One of the animals is decidedly a calculator; another likes to read or to explain figures; another detests reading but willingly taps out "spontaneous communications."

Without possessing much intrinsic probative value of its own, it is certain that all this fits in very badly with the supposition of a purely mechanical automatism operated by the person making the experiments. And on the other hand it bears a close analogy to the mediumistic "specialities"; that is, to the well-known fact that one "medium," for instance, is good for "physical effects" (i.e. gives rise around it to dynamic phenomena), but is not good for "psychography"; or produces "incarnations" but not "apports," etc. In the same way, typtology or rapping, more or less systematic, seems a fundamental gift, common to all the various kinds of "mediums." And the fact is perhaps of a certain value that precisely the same thing is true of "thinking" animals; although we must always remember that an analogous relation may only be apparent or extrinsic. Besides, the tone also of the "communications" in the two fields seems to me very much akin. I allude to the curious, angular, enigmatic, spasmodic, often playful and bantering communications, with frequent "unexpected replies" and philosophic platitudes. I find all these in Lola, and I remember similar stories of Rolf and of the horses, giving me an impression very like that which I get from the accounts of mediumistic seances "with intellectual effects."

Premising all this, we may suppose that a peculiar psychic concordance, which failing a better term might be called mediumistic, exists between Lola and her mistress. The mistress then in some way will have "communicated" through the dog the substance of her psychic self (perhaps with eventual autonomous additions from the canine or other psychic entity); all this happening, we must suppose, in a subliminal way, with partial psychical disassociation on the part of the authoress, if not also probably on the part of Lola, about which I am quite certain (and in this I agree with Neumann) that it absolutely does not understand anything or know anything of almost all the manifestations of thought which it exhibits.

There remain the questions (if the possibility of such duplicate mediumistic phenomena is admitted a priori to be possible) as to the point at which the normal relationship between a human person and an animal passes over into this supernormal one; and, finally, as to what particular known facts in the case of Lola, besides the rather too general analogies already mentioned, speak in favour of this hypothesis.

Into the mediumistic endowment of the investigator it seems to me useless to inquire since a priori many persons, so it seems, are more or less strikingly endowed, and the conditions which determine results are not sufficiently known. At the most there exist some indications—e.g. in Morselli's masterly work (2)—of the existence of some concordances between the phenomenology of mediumism and hysterical, hysteroid, or at least "sensitive" temperaments. And I believe that—with the help of their own publications, properly analysed—it would not be too difficult to attribute one or the other of such physio-psychic varieties to those persons who have up to the present obtained the best results with "thinking animals."

More interesting appears to me the investigation of the question whether animals themselves have already given any clear proof of being able to be "sensitive" in the mediumistic sense. And I must say that such a proof seems to have almost been reached.

I may refer on this subject to the exhaustive monograph published in 1905 by Bozzano (1) and written with the special competency and clearness that distinguish the well-known Genoese psychist.

Bozzano at that time was necessarily ignorant of the "thinking" animals, for it was only afterwards that they came to notice. But there were other authors who introduced the possibility (or the necessity) of a supernormal relationship in order to explain the Elberfeld facts, as soon as they were known. Perhaps the first in chronological order was De Vesme, who published in 1912 an interesting article in that sense (3), showing the many analogies between the phenomena of Elberfeld and mediumistic phenomena generally, e.g. the typtological particularities; the wrong orthography ("Firaz" tapped by the horse to express its own name "Zariff," "Dref" instead of "Ferd," etc.); solutions of difficult problems and invincible resistance to simple inquiries; immediate promptitude of correct replies to complicated mathematical problems, etc.