may be safely recommended. When an epidemic of cholera occurs in winter, then a relatively low state of the ground-water is found to prevail. In Munich three epidemics prevailed during the cold months: the first occurred between October and March, 1836-'37; the second from July to November, 1853-'54; and the third from July till April, 1873-'74. All three epidemics were associated with a relatively dry state of the earth, as was proved also by meteorological data concerning the rainfall. No investigation has been made as to the state of the ground-water during the period 1836 to 1854, but this investigation was first begun in 1856, so that for 1873 and 1874 the data were available; and it is only on the assumption that the condition of the soil as regards moisture was abnormal for the time of year that the long duration and strange division of the last epidemic could be accounted for. The subsoil-water sank from the end of June, 1873, till the beginning of August. On that occasion the germs of cholera probably came from Vienna, where the epidemic had prevailed since April. Two cases coming from Vienna, one in June, the other in July, could be vouched for. At the end of July the first illness from cholera occurred in Munich, but in individuals who had never come in direct contact with the infective cases. In every fresh outbreak, in 1836, in 1854, and in 1873, the same part (the northeast) of the town was the first to suffer. As the epidemics of 1854 and of 1873 developed at the same time of the year (the end of July), so by the middle of August the height of the epidemic was reached; it then fell off rapidly during September; during the whole of October only isolated cases occurred; and by the middle of November the epidemic had ceased in the higher lying parts of the town. It was thought that the disease had become extinct, and notwithstanding that it was considered strange that the summer epidemic had chiefly fallen upon the higher lying parts of the town, while the lower lying districts on this occasion had been altogether spared. In the middle of November, when the weather became colder, the epidemic reappeared, and attacked chiefly those lower lying districts which had escaped in the summer. It is impossible to trace the progress of contagion in time and space from one individual to another. The contagionists can not maintain that the unexpected falling off of cholera was due to the protection afforded by the previous prevalence of the disease, seeing that the lower lying districts had escaped the epidemic, and that the other inhabitants two months later suddenly lost their protection. Any one who studies the movements of the ground-water in Munich for the year in question will find that in the first half of August an event occurred which in suddenness and unexpectedness rivaled the retrogression of the epidemic in the second half of August. In the first half of this month there fell an abnormally large quantity of rain (one hundred and seventy-one millimetres), which excessive rainfall was the largest amount ever registered since the observations had been begun. The consequent wetness
