But when the hallucinatory character of collectively-perceived, or, as they may be styled for brevity, "collective" phantasms is recognised, there are difficulties of interpretation to be dealt with. On the telepathic hypothesis there are two modes in which a collective hallucination may be conceived to originate: (_a_) it may be communicated direct from a third person to each of the percipients; or (_b_) it may be communicated by telepathic infection from one percipient to another. The first explanation involves in most cases, as Mr. Gurney has pointed out (_Phantasms of the Living_, vol. ii. pp. 171, 172), serious theoretical difficulties. For on the view to which we are led by a review of all the evidence, a telepathic hallucination, like any other, is, as a rule, the work of the percipient's mind, and is not transferred ready made from the agent. As such it is frequently of slow growth, and there are grounds for believing that it is sometimes not externalised for the percipient's senses until some hours after the receipt of the original telepathic impulse. We should hardly expect, therefore, to find two percipients independently developing similar hallucinations, and at the same moment. But in most of the cases of collective hallucination hitherto reported, the hallucinations have been, so far as could be ascertained, similar and simultaneous, so as indeed to suggest a real figure rather than a hallucination. Moreover, in well-attested recent narratives it rarely happens that a connection between the hallucination and any unusual state of the person represented is clearly established; whilst in many, perhaps most cases, the hallucination has not been recognised as resembling any person known to either percipient, and has in some instances been purely grotesque. In most cases, therefore, it seems easier to believe that we have to deal with a contagious hallucination, which, whether initiated by a telepathic impulse, or purely subjective in its origin, has been transferred telepathically from the original percipient to others in his company at the time. In some cases, indeed, it is no doubt permissible, as suggested by Mr. Gurney, to conjecture that the minds of all the percipients may have been directly influenced by the agent, and that subsequently an overflow from the mind of one of the percipients may have served to reinforce the original impulse, and determine the exact moment of the explosion in his co-percipients, just as the current regulates the exact hour of striking in electrically synchronised clocks. Or again, the mind of each percipient may react upon the others. There are, however, a few cases where the percipients appear to have had experiences relating to the same event neither precisely similar nor simultaneous, which seem to require the hypothesis of an impulse in each case directly derived from the person represented. Some cases of the kind are given in _Phantasms of the Living_ (vol. i. p. 362; vol. ii.