belongs, the probability is great that the weight of evidence has been found altogether overwhelming. Let us now, however, turn to cases in which, while many observations seem to point to some result, it has appeared that, after all, those observations must have been illusory. A striking instance in point is found in the perplexing history of the supposed satellite of Venus. On January 25, 1672, the celebrated astronomer, J.D. Cassini saw a crescent shaped and posited like Venus, but smaller, on the western side of the planet. More than fourteen years later, he saw a crescent east of the planet. The object continued visible in the latter case for half an hour, when the approach of daylight obliterated the planet and this phantom moon from view. The apparent distance of the moon from Venus was in both cases small, viz., only one diameter of the planet in the former case, and only three-fifths of that diameter in the latter. Next, on October 23, 1740, old style, the optician Short, who had had considerable experience in observation, saw a small star perfectly defined but less luminous than Venus, at a distance from the planet equal to about one-third of the apparent diameter of our moon. This is a long distance, and would correspond to a distance from Venus certainly not less than the moon's distance from the earth. Short was aware of the risk of optical illusion in such matters, and therefore observed Venus with a second telescope; he also used four eye-pieces of different magnifying power. He says that Venus was very distinct, the air very pure, insomuch that he was able to use a power of 240. The seeming moon had a diameter less than a third of Venus's, and showed the same phase as the planet. Its disc was exceedingly well defined. He observed it several times during a period of about one hour.