and diminishes their nutritive properties. To obviate this, we must find means to sterilize the liquids without heating them. Filtration has been employed, and a number of adaptations of apparatus have been devised by means of which this object is accomplished satisfactorily.
In order to study the germs in the air, we must not only possess a sensitive and wholly pure liquid, but must have means of arranging it for the cultivation of the organisms, under such circumstances that we may be sure it shall contain no germs except those that are derived from the air we introduce for the experiment. MM. Chamberland and Miquel have employed simple apparatus which seem to effect this purpose perfectly.
Fig. 6.—Micrococcus bacilliformis (after M. Miguel). A, adult plant; B, examples of hypertrophied cells; C, chain at maturity; D, chain destroyed. Magnified 1,000 diameters.
If the experiments are made with rain-water, to ascertain the number of germs it collects in passing through the air, it may sometimes happen that, when a determined volume of water is evenly distributed in a considerable number of the cultivation-tubes, only a part of the tubes will become troubled. Generally, it may be said that if the water contains as the average one bacterium per cubic centimetre,
