The proportionate number of deaths at all ages from diarrhoea corresponds pretty closely with those of infants.  To prove this, I made calculations for three years, and ascertained that only 3.9 per cent. of all the deaths from this disease were registered in the weeks having a temperature of less than 50 deg.; 11.9 per cent. in the weeks having a temperature between 50 deg. and 60 deg.; while in the comparatively few weeks in which the temperature exceeded 60 deg.  F., as many as 84.2 per cent. of the total number of deaths was registered.  In the sixteen years, 1840-56, for which many years ago I made a special inquiry, only 18.9 per cent. of all the deaths from diarrhoea occurred in winter and spring, against 81.1 per cent. in summer and autumn.  In the twenty years, 1860-79, there were seven years in which the summer temperature was in defect when the mortality per 100,000 inhabitants of London was 200; while in ten summers, during which the temperature was in excess by 2 deg. or less, the mortality was 317 per 100,000.  The mean temperature was largely in excess, that is to say, more than 2 deg. plus in three of these summers, when the mortality reached 339 per 100,000 inhabitants.

These figures show that great care should be taken in hot weather to prevent diarrhoea, especially among young children; by frequent washing with soap and water to insure cleanliness, and proper action of the skin; by great attention to the food, especially of infants fed from the bottle; free ventilation of living rooms, and especially of bedrooms; and by protection, as far as possible, being afforded from a hot sun, as well as by avoiding excessive exercise.  All animal and vegetable matter should be removed from the vicinity of dwelling-houses as quickly as possible (indeed, these should be burnt instead of being put in the dust-bin), the drains should be frequently disinfected and well flushed out, especially when the mean daily temperature of the air is above 60 deg.  F.

Time will not admit of more than a mere mention of the relations between meteorological phenomena and the mortality from many other diseases and affections, such as apoplexy from heat, sunstroke, liver diseases, yellow fever, cholera, whooping-cough, measles, etc., especially as the state of our knowledge on the subject is so very limited.  A comparison between the mortality from several diseases in this and other countries shows that certain of these do not prevail under closely corresponding conditions.  Thus the curves of mortality from whooping-cough, typhoid fever, and scarlet fever do not correspond with the curves of temperature in both London and New York, and the same may be said of diarrhoea in India.  It is therefore evident that some other cause or causes than a varying temperature must be concerned in the production of an increased death-rate from these diseases.  The subject is of great importance, and I do not despair of our obtaining some day a knowledge of the agents through which meteorological phenomena act in the production of increased and decreased death rates from certain diseases, and the means by which, to a certain extent, these injurious effects on man may be presented.