Mr. Edward Clodd is the author of several handbooks of science--'The Story of Creation,' 'A Manual of Evolution,' and others. Now, in a signed review of a book, a critique published in 'The Sketch' (October 13, 1897), Mr. Clodd wrote about the Census: 'Thousands of persons were asked whether they had ever seen apparitions, and out of these some hundreds, mostly unintelligent foreigners, replied in the affirmative. Some eight or ten of the number--envied mortals--had seen "angels," but the majority, like the American in the mongoose story, had seen only "snakes."... In weighing evidence we have to take into account the competency as well as the integrity of the witnesses.' Mr. Clodd has most frankly and good-humouredly acknowledged the erroneousness of his remark. Otherwise we might ask: Does Mr. Clodd prefer to be considered not 'competent' or not 'veracious'? He cannot be both on this occasion, for his signed and published remarks were absolutely inaccurate. First, thousands of persons were _not_ asked 'whether they had seen apparitions.' They were asked: 'Have you ever, when believing yourself to be perfectly awake, had a vivid impression of seeing, or being touched by a living being or inanimate object, or of hearing a voice; which impression, so far as you could discover, was not due to any external physical cause?' Secondly, it is not the fact that 'some hundreds, _mostly unintelligent foreigners,_ replied in the affirmative.' Of English-speaking men and women, 1,499 answered the question quoted above in the affirmative. Of foreigners (naturally 'unintelligent'), 185 returned affirmative answers. Thirdly, when Mr. Clodd says, 'The majority had seen only "snakes,"' it is not easy to know what precise sense 'snakes' bears in the terminology of popular science. If Mr. Clodd means, by 'snakes,' fantastic hallucinations of animals, these amounted to 25, as against 830 representing human forms of persons recognised, unrecognised, living or dead. But, if by 'snakes' Mr. Clodd means purely subjective hallucinations, not known to coincide with any event--and this _is_ his meaning--his statement agrees with that of the Census. His observations, of course, were purely accidental errors. The number of hallucinations representing living or dying recognised persons in the answers received, was 352. Of first-hand cases, in which coincidence of the hallucination with the death of the person apparently seen was affirmed, there were 80, of which 26 are given. The non-coincidental hallucinations were multiplied by four, to allow for forgetfulness of 'misses.' The results being compared, it was decided that the hallucinations collected coincided with death 440 more often than ought to be the case by the law of probabilities. Therefore there was proof, or presumption, in favour of some relation of cause and effect between A's death and B's hallucination.