The water of small pools and marshes should be avoided, and even springs should be regarded with suspicion, unless some idea can be formed as to the origin of the water. If this be from a deep source, it may of course, be safely used, but care must be taken not to mistake surface drainage that has oozed a short distance under ground for a true deep spring. Caution in this matter is especially necessary in hill country; and in doubtful cases it is well to get water analysed by an expert before adopting it as a permanent supply, for it is noteworthy that some springs of the highest reputation have been shown to be extremely impure, the sparkle of their waters being really due to their being charged with the gases of decomposition. Not unfrequently, however, the tropical sojourner has no choice as to his water supply, and must make the best of perhaps a very bad source. Under such circumstances, all water used for drinking, or in the preparation of food, must be specially treated so as to remove or destroy any of the germs of disease which it may contain. It is desirable, though very difficult, to treat water used for bathing in the same way, but as a rule, one has to be content with taking every precaution against such water entering the mouth. Ordinary filters, it must be clearly understood, are not only useless, but even worse; for the moist filtering agent, clogged as it is with the coarse organic and inorganic _débris_ that it has strained out of the water, is quite capable of acting as a cultivating medium for microbes, on which they can multiply so enormously that however clear the water may appear to the eye when it issues, its really dangerous impurity, so far from being diminished, has been enormously increased. There is only one form of filter that can be trusted to remove the minute organisms that are the active agents in the propagation of disease, and that is that in which the water has to pass through a piece of biscuit porcelain, the pores of which are so excessively minute that even the smallest of the bacteria are excluded. The employment of such filters on a domestic scale is, however, extremely difficult, for they naturally act so slowly that a very large appliance must be used to secure an adequate supply. Added to this, their efficiency depends on the perfection of a number of rubber connections, a material which deteriorates very rapidly in hot climates, and except by the increased rapidity of flow, it is not very easy to detect the fault. In any case, they should never be used except under constant personal care, and one way and another, they require a good deal of attention; besides which they are heavy, and quite unsuitable for camp or travelling. On the other hand, boiling for a few minutes gives a security quite sufficient for practical purposes, and requires no more formidable appliances than an ordinary kettle or saucepan, which are available everywhere. I am