Dysentery arises in a very similar way, and is, no doubt, merely an accentuated form of the disease with a tendency to become chronic, incidental to a tropical climate. The effect of chilling of the body, on which so much stress has been laid, is probably to increase the susceptibility of the system to the entrance of the poison from without. Attacks of dysenteric diarrhoea with discharges of blood and mucus per rectum, are occasionally associated with outbreaks of diarrhoea in this country.

Although it is unquestionable that dysentery and acute diarrhoea in the vast majority of cases appear to arise de novo, independently of the contagion of a previous case, yet it is also as certain that the diarrhoeal evacuations help to spread the disease in certain cases. It is not impossible that the diseased process may originate a specific poison within the body, which on evacuation has an infective virulence not inferior to the enteric fever or cholera poison.

From the seasonal curve for diarrhoea (fig. 72) it will be seen that the mortality begins to increase about the middle of June, rises rapidly to its maximum at the end of July or early in August, and falls somewhat less rapidly throughout August, September, and October.

A most exhaustive and excellent report by Dr. Ballard upon the "Causation of the annual mortality from diarrhoea, which is observed principally in the summer season of the year" has been recently issued. The following is a very brief epitome of Dr. Ballard's observations.

The summer rise of diarrhoeal mortality in the large towns does not commence until the mean temperature

* Supplement to Report of Medical Officer of the Local Government Board for 1887.

recorded by the earth thermometer, placed 4 feet below the surface, has attained somewhere about 56° F.-no matter what may have been the temperature previously attained by the atmosphere, or recorded by the 1-foot earth thermometer. The maximum diarrhoea mortality of the year is usually observed in the week when the 4-foot earth thermometer attains its mean weekly maximum. The diarrhoea mortality declines with the 4-foot earth thermometer; and this decline takes place very much more slowly than that of the atmospheric temperature or of the 1-foot earth thermometer, so that the mortality from epidemic diarrhoea may continue long after the air temperature has fallen, even into the 4th quarter of the year.

The soils most favourable to a high diarrhoea mortality are those of sand, gravel, or marl, in which the constituent particles are small but freely permeable by air and water, and which contain organic matters of animal origin from "made ground," from manured surfaces, or from soakage of excretal refuse from privies, cesspools, and sewers. The soil must be moist, but the moisture must not be sufficient to preclude the free admission of air between the interstices, e.g., soils in which the subsoil water stands sufficiently near the surface to maintain by capillary attraction the dampness brought about by previously greater nearness of the water to the surface, or marly soils containing clay sufficient to imprison enough of the water saturating it at some time previously. The moisture of the soil may arise from surface water sinking into the earth around houses, as well as from the subsoil water from below.

Other factors conducive to a high diarrhoea mortality. are crowding of houses on area, so that they have deficient light and external ventilation, the building of

houses back-to-back, and the keeping of milk and other foods in underground cellars exposed to telluric emanations, or in pantries liable to the entry of drain or sewer air.

As previously stated, the disease is mainly one of early childhood (0-5 years), but its incidence is by far the greatest on hand-fed infants. The attacks are

usually extremely sudden in their onset; and that diarrhoea is merely one symptom or feature of the illness, is shown by the fact that many of the organs of those who have succumbed are found to be highly degenerated, more especially the kidneys, the liver (fatty degeneration), and the spleen. The lungs too are often the seat of pneumonic inflammation. Dr. Klein has failed to find anything in the tissues, blood, or excreta of these cases, to indicate that the malady is due to a microbe developing within the alimentary canal or permeating any of the tissues. But in certain groups of cases of epidemic diarrhoea, the disease is apparently communicable from person to person by means of the exhalations from the stools, and Dr. Ballard believes that in the excretions of such cases a specific micro-organism may possibly be found.

The following provisional explanation of the occurrence of epidemic diarrhoea is offered by Dr. Ballard:"That the essential cause of diarrhoea resides ordinarily in the superficial layers of the earth, where it is intimately associated with the life processes of some micro-organism, not yet detected, captured, or isolated.

That the vital manifestations of such organism are dependent, among other things, perhaps principally upon conditions of season, and on the presence of dead organic matter, which is its pabulum.

That, on occasion, such micro-organism is capable

of getting abroad from its primary habitat, the earth, and having become air-borne, obtains opportunity for fastening on non-living organic material, and of using such organic material both as nidus and as pabulum in undergoing various phases in its life history.

That in food, inside of, as well as outside of the human body, such micro-organism finds, especially at certain seasons, nidus and pabulum convenient for its development, multiplication, or evolution.

That from food, as also from the contained organic matter of particular soils, such micro-organisms can manufacture, by the chemical changes wrought therein through certain of their life processes, a substance which is a virulent chemical poison.*

That this chemical substance is, in the human body, the material cause of epidemic diarrhoea."

During the decennium 1871-80, the death-rate in England and Wales from diarrhoeal diseases was 0.93 per 1000 living at all ages. Under 5 years of age, the death-rate was 5'7 per 1000; and although this high rate is largely contributed to by the improper nourishment and feeding of infants, there can be no doubt that insanitary conditions, of the kinds named above, play a large part in its production. For the 8 years 1881-8, the average death-rate in England from diarrhoeal diseases was 0.76 per 1000 living at all ages.

Rabies.

Rabies is a specific disease of which the virus-probably a microbe-has not been known to be transmitted otherwise than by inoculation, i.e., by the bites of rabid * Probably one of the alkaloidal Ptomaïnes.