The headstones, it will be remembered, carry us back only a few centuries. But, so far as these memorials reach, they tell overwhelmingly in favour of the superstition. Twenty years ago, this evidence was much more patent to the eye. A very general exception, however, must be made of the town churchyards, in which the old tombstones are often found evenly distributed over the available space. Several reasons may be adduced for this non-observance of the usual practice. The town church, as a rule, stands centrally in its graveyard; it is commonly approached by several alleys and paths, leading from a thoroughfare or market-place; it is frequently cruciform in plan, with North and South doors, and this design, wherever met with, seems to have had some effect in counteracting the fears of the ignorant. Above all these reasons, must be set the fact that populous districts would soonest lose touch with the superstition, so that, even before the era of headstones, the belief retained but a comparatively feeble foothold among the inhabitants of towns. The argument from monuments fails in cases such as these. We must seek the tradition in the rural districts, and turn back to a time when it was more firmly held. We shall then see that the present exceptions, numerous as they are, cannot invalidate the general practice. They represent what Professor L. C. Miall, writing on a vexed question in botany, calls “negative exceptions.” I venture to repeat his witty illustration: “A wooden leg is used to enable a man to walk when he has lost his natural leg. If you saw a one-legged man walking with a pair of crutches, and no wooden leg at all, would that shake your belief in the motive for using wooden legs[902]?” The “negative exceptions” which we have been studying may, indeed, testify to a weakening of tradition, or occasionally, to its apparent local non-existence, but they do not abolish it, or change its purport one whit. How it chances that, of two adjacent parishes, the present-day evidence shows the belief at work in the first, but seemingly unknown, or in abeyance, in the second, is a matter upon which we can only speculate. The puzzle reminds one of the difficulty which meets the palaeontologist when he strives to explain why one line of animal descent stops, and remains fixed, while another continues to develop; why creatures which seem to be completely adapted to their surroundings become extinct; why one species or genus is taken, and another left.