Again, if the ice field is wind-swept by air bearing an unduly rich complement of bacteria, as may be expected in the vicinity of populous cities, for example, then the ice will reflect in its bacterial contents the undesirable neighbourhood in which it was produced. Water in repose, again, yields purer ice than water in movement during freezing, for during rest opportunity is given for the bacteria present in suspension to subside, the process of sedimentation or deposition of bacteria which takes place under these conditions playing an important part in water-purification; when, however, the water is disturbed by swift currents, or agitated by storms, this process is interrupted, and the bacteria become entangled in the ice and frozen _in situ_. The importance attaching to the physical conditions under which ice is produced in enabling an estimate to be formed of the safety or otherwise of the same for consumption may be gathered from the following extract from an American report on the subject:-- "On the whole it is evident that the conditions surrounding water when it freezes are very important factors in determining the purity of the ice formed. If there is a considerable depth of water in portions of a somewhat polluted pond or river, and the ice is formed in these portions in comparatively quiet water with but little matter in suspension, this ice will probably be entirely satisfactory for domestic use. On the other hand, ice formed in shallow portions of such ponds or rivers, even during still weather, or in any portion if there is a considerable movement of the water by currents or wind while it is forming, may be rendered by these conditions entirely unfit for domestic use." We have learnt that ice contains bacteria, that its bacterial contents are to a certain extent dependent upon the bacterial quality of the water before crystallisation, and that an important factor in determining its purity is afforded by the physical conditions prevailing at the time of freezing. It will be of interest to ascertain in more detail what effect the process of freezing has upon the number of bacteria present in the water--what is the degree of bacterial purification effected during the conversion of water into ice. Now Professor Uffreduzzi, in his investigations on the ice-supply of Turin, part of which is derived from a much-polluted portion of the River Dora, found that about 90 per cent. less bacteria were present in the ice than were present in the water from which it was produced. In the making of ice, therefore, a remarkable removal of bacteria may be effected which approaches very nearly the degree of bacterial purification which is achieved during the best-conducted sand-filtration of water.