Whether the S.P.R. will ever succeed in bottling a ghost, and in submitting it to the tests necessary to convince science, matters little. The real fruit of its labours will be to "convince men of sin," to convict science of being unscientific, and criticism of being uncritical--of being biassed by fashion to the extent of refusing to examine evidence which must be either admitted or explained away. Scepticism and credulity alike are hostile both to science and religion, and it is the common interest of these latter to secure a full recognition, on the one side of the principle of faith, that with God all things are possible; and on the other, of the principle of science which is: to prove all things, and hold fast that which is good. Credulity tends to make the actual co-extensive with the possible; while scepticism would limit the possible to the known actual. The true mind would be one in which faith and criticism were so tempered as to secure width without slovenliness, and exactitude without narrowness. II. How, apart from the imperfect lingering tradition of some primitive revelation, the belief in a surviving soul originates with contemporary savages, or might have originated among still ruder past races, is a question of some interest, not only for its own sake, but for the sake of whatever little light it may throw upon the more vital question as to the value of that belief. Had the doctrine of souls no other origin than a false inference from the ordinary phenomena of sleeping and dreaming; were it in no sense an instinctive belief, suggested perhaps and confirmed by supernormal facts, it would still have interest for the anthropologist as one of those almost necessary and universal errors through which the human mind struggles to the truth, such as the errors of astrology or alchemy; but it would in no way contribute to the argument for immortality _ex consensu hominum_--an argument of much avail when it is a case of man's instinctive judgments and primary intuitions, which are God-given, but of ever less value in proportion as there is a question of deductions, inferences, and self-formed judgments. Even if we discard the dream-theory altogether, we get no support from the consensus of savages as to the soul's survival, unless we have reason to think that the facts on which their inference rests are truly, and not only apparently, supernormal, and are, moreover, such as leave no other inference possible.