The collection of samples, though it appears simple enough, is sometimes a difficult and responsible undertaking. Complicated apparatus is rarely necessary, and fallacies will generally be avoided by observing two directions. In the first place, the sample should be chosen as representative as possible of the real substance or conditions we wish to examine. Some authorities advise that it is necessary to allow the tap to run for some minutes previous to collecting the sample; but if we desire to examine for lead chemically or for micro-organisms in the pipes biologically, then such a proceeding would be injudicious.[11] Hence we must use common sense in the selection and obtaining of a sample, following this one guide, namely, to collect as nearly as possible a sample of the exact water the quality of which it is desired to learn. In the second place, we must observe strict bacteriological cleanliness in all our manipulations. This means that we must use only sterilised vessels or flasks for collecting the sample, and in the manipulation required we must be extremely careful to avoid any pollution of air or any addition to the organisms of the water from unsterilised apparatus. A flask polluted in only the most infinitesimal degree will entirely vitiate all results. Accompanying the sample should be a more or less full statement of its source. There can be no doubt that, in addition to a chemical and bacteriological report of a water, there should also be made a careful examination of its source. This may appear to take the bacteriologist far afield, and in point of fact, as regards distance, this may be so. But until he has seen for himself what "the gathering-ground" is like, and from what sources come the feeding streams, he cannot judge the water as fairly as he should be able to do. The configuration of the gathering-ground, its subsoil, its geology, its rainfall, its relation to the slopes which it drains, the nature of its surface, the course of its feeders, and the absence or presence of cultivated areas, of roads, of houses, of farms, of human traffic, of cattle and sheep--all these points must be noted, and their influence, direct or indirect, upon the water carefully borne in mind. When the sample has been duly collected, sealed, and a label affixed bearing the date, time, and conditions of collection and full address, it should be transmitted with the least possible delay to the laboratory. Frequently it is desirable to pack the bottles in a small ice case for transit. On receipt of such a sample of water the examination must be immediately proceeded with, in order to avoid, as far as possible, the fallacies arising from the rapid multiplication of germs. Even in almost pure water, at the ordinary temperature of a room, Frankland found organisms multiplied as follows:-- No. of Germs Hours. per cc. 0 1,073 6 6,028 24 7,262 48 48,100