In experiments conducted as we have just described, the yeast which appears, and which soon sets up an active alcoholic fermentation, is introduced in the first place by atmospheric air, from which germs are constantly falling either upon the film of mycoderma or upon the objects that are employed in the successive manipulations. Two peculiarities in these experiments first opened our eyes to the existence of this cause of error. We sometimes found at the bottom of the flasks in which we had submerged the efflorescence, along with the cells of mycoderma, large, spherical cells of _mucor mucedo_ or _racemosus_, ferment-cells that we shall soon learn to recognize in studying this curious fungoid growth. The existence of _mucor mucedo_ or _racemosus_, where we had only sown _mycoderma vini_, was to us a proof that one or more spores of that mucor had been introduced by the surrounding air. If then, we reasoned, the air can introduce spores of mucor into our field of operations, why should it not introduce cells of yeast, especially in our laboratory? Again, it sometimes happened that a negative result was obtained. Harassed by doubts about the reality of this transformation, which accorded so well with, the physiological theory of fermentation we had been led to adopt, we repeated the experiments many times, and in some cases we failed to detect any appearance whatever of a transformation of mycoderma into yeast cells, although the conditions under which each of the experiments was conducted had been as similar as could be. We were at a loss to account for this inactivity in the cells of the mycoderma. Even in the most favourable cases of the supposed fermentation, it was evident that a host of cells of _mycoderma vini_ did not become cells of yeast; but how could it possibly be admitted that amongst the millions of submerged cells, none were adapted for transformation, if that transformation were at all possible?