It is very remarkable that some of the _torulæ_ in Fig. 12 are not ferments; they do not cause sugar to decompose into alcohol and carbonic acid, any more than the _mycoderma vini_ does; but, nevertheless, there may be an absolute similarity in aspect, development, shape, and size between the alcoholic ferment, properly so called, and these _torulaceæ_. We must here cite a case in proof. On May 28th, 1872, in one of the rooms of our laboratory, we broke the fine points of a series of flasks, similar to those used in our previous experiments, containing must of grapes and deprived of air. We then closed up the ends immediately after the sudden entrance of the exterior air. One of these flasks developed only one kind of organism, which belonged to the _torulaceæ_. On June 7th this was sufficiently abundant to cover all the sides with a white deposit, and the surface of the liquid appeared quite free of any _mycoderma vini_. To assure ourselves that we were actually dealing with one kind of _torulæ_, unattended by fungoid growths, we waited until June 14th, but the aspect of things remained unchanged. On that day we opened the flask; there was no escape of gas to indicate that the interior pressure was greater than the exterior. We then subjected the plant to microscopic examination. It was quite homogeneous, and formed of a mass of cells, absolutely identical in aspect and size with old cells of ordinary yeast (Fig. 13). [Illustration: Fig. 13.] We distilled all the liquid, of which there was about 100 c.c., (3-½ fl. oz.) without obtaining any trace of alcohol in our distillate; we collected 33 c.c. (about an ounce) of the distillate, which we distilled again, and we submitted the distillate a third time to distillation; but even then there were no more signs of the presence of alcohol than there had been in our first distillate.[46] We may safely conclude that our _torula_, in the course of its development in must, with a weight that would have been very appreciable, did not produce, by its action of multiplication 1/10,000 of 1 c.c. of alcohol. Under the following conditions we obtained a slightly different result, which, nevertheless, confirmed the preceding one.