to depict more strikingly the behavior of these organisms toward the culture substance-in this case apple juice. The weighings made at stated intervals show the destruction of sugar through the loss. suffered by reason of the carbon dioxid given off. The weights were taken each twenty-four hours and the results noted on the record sheets. Fig. 4 depicts in a striking manner the daily loss of carbon dioxid by the height to which the line designating the activity of each yeast rises above the base. The chart is in fact a picture of the vital activities of each yeast organism in the test.

So far as the weighings showed no change occurred during the first twenty-four hours, but it is a fact, frequently verified by the writer by microscopic examinations, that there results a large increase in the number of yeast cells during the first day, but until the liquor is fairly saturated with the carbon dioxid gas none is given off, and therefore the destruction of sugar would not be shown by weighing. The second day all the tests showed some loss, reaching 0.75 of a gram in the case of the mixed culture. This greater activity was caused by the larger amount of yeast sown. The next most active yeast is No. 100. This is always the case when S. apiculatus is compared with other yeasts, and it is this precocity of development that renders it so obnoxious in the fermentation industries. In its final development No. 100 makes a poor showing compared with the other yeasts. Its maximum occurs on the fourth day at 2.05 grams, and on the same day No. 37 reaches its maximum at 6.33 grams loss of carbon dioxid, though it was one of the slowest to start.

Between these extremes lie all the other yeasts used, the mixed culture showing its climax at about half the height of No. 37. An interesting observation, which holds good in other tests, is that a mixture. of several pure yeasts has not shown the same power of fermentation that many of the same yeasts show when sown alone. In this particular case, and in fact whenever S. apiculatus is used, this result, to a considerable extent, should probably be credited to the fact that this undesirable form is for a time dominant in the liquor to the partial exclusion of more desirable ferments.

MISCELLANEOUS ANALYSES OF FERMENTED MUSTS.

During the several years that the tests of pure yeasts and organisms associated with them have been in progress many chemical analyses of the fermented liquor have been made, and a number of these analyses are presented in Table IV for reference. Generally the yeasts do not show striking differences in the amount of alcohol produced when used in the same must. There are, however, some interesting differences, and these are brought out by the tabulated analyses. This work was undertaken to collect preliminary data on

the activity of yeasts as chemical agents, and complete check tests were not made in many cases. A considerable number of check tests and analyses were made, however, and the data herein presented represent fairly the results obtained.

The results of the analyses of must inoculated with several of the fungi associated with the yeasts are interesting. With Penicillium in one case (test 82) the destruction of 2.12 per cent of sugar is shown and only 0.74 per cent of alcohol was formed, and in the other 2.14 per cent of sugar was consumed and 1.46 per cent of alcohol formed, relatively a very high result. But the species of Aspergillus used in three long tests appeared to be incapable of forming alcohol except in one instance, where 0.28 per cent was produced by this fungus. Yet these fungi destroyed 2 per cent and over of sugar. Evidently from these results it is correct to consider these organisms disturbing agencies, especially where fermentation proceeds slowly.

The forms of Torula used from two different sources in tests Nos. 88 and 107 show decided strength as alcoholic ferments. The two cultures used were quite different as to characteristics of the cells, so much so as to indicate difference in species. The fact that this organism forms alcohol so readily renders it all the more objectionable when associated with yeasts in the fermentation industries, because during growth its secretions are likely to injure the bouquet of the product.

The three forms, or presumably species, of Mycoderma used show striking differences in their fermenting power, varying from less than one-half of 1 per cent to more than 4 per cent of alcohol. All of these inverted sucrose, No. 92 practically all of it, and consumed 11.46 per cent of sugar to form 4.15 per cent of alcohol, a poor result in that respect. In the case of this strong growing form it appears that it might become a decided competitor of yeasts, especially under conditions where fermentation progresses slowly. This fungus is also a very undesirable ferment. Dematium did not show much power to form alcohol so far as tested.