It covers the whole field of automatism (as in automatic writing) perhaps of the divining rod, certainly of crystal visions and of occasional hallucinations, as Mr. Galton, in this last case, expressly declares. Psychologists at least need not be told that such faculties cannot, any more than other human faculties, be always evoked for study and experiment. Our evidence for these faculties and experiences, then, is usually of the class on which the psychologist relies. But, when the psychologist, following Leibnitz, Sir William Hamilton, and Kant, discusses the Subconscious (for example, knowledge, often complex and abundant, unconsciously acquired) we demonstrated by examples that the psychologist will contentedly repose on evidence which is not evidence at all. He will swallow an undated, unlocalised legend of Coleridge, reaching Coleridge on the testimony of rumour, and told at least twenty years after the unverified occurrences. Nay, the psychologist will never dream of procuring contemporary evidence for such a monstrous statement as that an ignorant German wench unconsciously acquired and afterwards subconsciously reproduced huge cantles of dead languages, by virtue of having casually heard a former master recite or read aloud from Hebrew and Greek books. This legend do psychologists accept on no evidence at all, because it illustrates a theory which is, doubtless, a very good theory, though, in this case, carried to an extent 'imagination boggles at.' Here the psychologist may reply that much less evidence will content him for a fact to which he possesses, at least, analogies in accredited experience, than for a fact (say telepathic crystal-gazing) to which _he_ knows, in experience, nothing analogous. Thus, for the mythical German handmaid, he has the analogy of languages learned in childhood, or passages got up by rote, being forgotten and brought back to ordinary conscious memory, or delirious memory, during an illness, or shortly before death. Strong in these analogies, the psychologist will venture to accept a case of language _not_ learned, but reproduced in delirious memory, on no evidence at all. But, not possessing analogies for telepathic crystal-gazing, he will probably decline to examine ours. I would first draw his attention to the difference between revived memory of a language once known (Breton and Welsh in known examples), or learned by rote (as Greek, in an anecdote of Goethe's), and verbal reproduction of a language _not_ known or learned by rote but overheard--each passage probably but once--as somebody recited fragments. In this instance (that of the mythical maid) 'the difficulty ... is that the original impressions had not the strength--that is, the distinctness--of the reproduction. An unknown language overheard is a mere sound....'[6]