There is not, on the surface, much uniformity in the weather preceding the epidemics of the sweat in 1508, 1517, 1528 and 1551. In the first of these the winter was mild and the early summer excessively hot and dry; in the second the winter and spring were remarkable for drought, with several weeks of intense black frost in the middle period; in the remaining two the antecedent appears to have been an excessive rainfall. But in all the four we shall find that the law of the sub-soil water, as formulated by the recent Munich school with reference to epidemic outbursts, was exemplified. According to that law, the dangerous products of fermentation arise from the soil when the pores of the ground are either getting filled with water after having been long filled with air, or are getting filled with air after having been long filled with water. It is the range of fluctuation in the ground-water, either downwards or upwards, that determines the risk to health; and in two of the years of the sweat, 1508 and 1517, we find that there had been a rise from a very low level of the wells, while in the other two, 1528 and 1551, the wells had begun to fall after standing for a length of time at an unusually high level. If this reading of the somewhat imperfect data can be trusted, it is at one and the same time an explanation of the outbreak of the sweat in the respective seasons, and a confirmation of the hypothesis that the virus of the sweat had its habitat in the ground. That hypothesis is, indeed, supported by so great a convergence of probabilities, both for the English sweat and for the endemic sweat of France[540], that it may be used to explain the seasonal incidence without laying the argument open to the charge of running in a vicious circle. Whatever had been the kind of weather determining the successive outbreaks of the sweat, it is clear that the favouring circumstances were in general not the same as those of the bubo-plague. The greater outbursts of plague, as we shall see, were in 1500, 1509, 1513, 1531, 1535, 1543, 1547, and other years not sweat-years. It is only in the autumn of 1517 that the plague overlaps somewhat on the sweat, and even then it becomes noticeable mostly in the winter following the decline of the sweat. The two poisons had existed in English soil side by side, but had not come out at the same seasons; also the sweat had been mostly a disease of the greater houses, and the plague mostly of the poorer. The Extinction of the Sweat in England.