ods be devised are the ones that are still used to detect the presence of these plants in the atmosphere. Generally, these methods are founded on the fact that organic liquids become peopled with microphytes, or remain unchanged, according as the air comes to them charged with its normal quantity of germs, or after having been cleared of them by filtration. We can, therefore, examine the bacteria in the atmosphere by causing the air or the rain-water containing them to pass into liquors favorable to their nutrition, but previously free from them. Liquids for the culture of bacteria are easy to procure. Among them are the mineral solutions of Pasteur and of Cohn, infusions of hay and of turnips, neutral urine, broths of chicken or beef, and Liebig's extract. It is, however, very hard to obtain such liquids absolutely pure of every living being. Eminent physiologists have thought that all the germs could be killed by boiling them for a considerable time. Apparently the protoplasm, being an albuminoid substance, would be coagulated at a temperature of between 167° and 176°; but very exact experiments have shown that while the protoplasms of different living beings belong to the same class of substances, they are not identical. Marked differences in this respect have been perceived even in the same beings. Thus, in the bacillus the protoplasm of the developed organism and that of the spore are not of the same quality. The former is in active life, the second in a state of life so low that it appears latent. A spore of this kind, as M. Chamberland has observed, will resist boiling water for hours, while the bâtonnet which is developed from it would perish rapidly in the same water at 122°.
M. Koch has conceived a method of discontinuous heating to sterilize liquids that are coagulable by heat. He raises his liquid to a temperature of not quite 158° to kill the adult bacteria; then having cooled it, to give the spores time to germinate, he raises it again to about 158°; and he believes that he can in this way destroy all the germs. M. Miquel makes a just criticism of this singular theory. We arrange that the spores of the microbes "must germinate in twelve or twenty-four hours, so that we may surely kill them if they go into the trap we set for them. But some of the germs may be obstinate or hardy, and we make a new trial, and for prudence a third and a fourth trial, after which we assume that all the bacteria have been destroyed. Unfortunately for the method of discontinuous heating, there are wary germs the development of which does not begin till after the fifth, the tenth, and even the twentieth day, and which, far from being stimulated in their growth by the successive heatings, at every repetition shut themselves up more closely in their latent seed-life. This method of sterilization can not, then, be depended upon."
A still more subtile cause of error must be guarded against. Cohn's mineral liquid will remain clear for an indefinite period after having
