To obtain a complete conversion of the starch into sugar and alcohol is impracticable. Even under the best conditions the yield of alcohol is, from various causes, appreciably lower than the quantity theoretically obtainable if the starch could be all converted into alcohol and carbon dioxide only. This theoretical quantity would be: -

About six-sevenths of the theoretical yield is obtained. Taking starch as the starting-point, it is estimated that from 12 to 20 per cent. is generally lost in one way or other by the time the fermentation is finished. From 6 to 10 per cent. of the starch and dextrin remains unfermented. During the fermentation, 2 to 3 per cent. of glycerol is formed; a part of the sugar is used up in providing nutriment for the yeast; and a little alcohol is lost by evaporation. Any side-fermentations that occur through the agency of bacteria or other non-yeast organisms will still further diminish the yield of alcohol. In a badly-conducted operation the alcohol produced may represent not more than about 72 per cent. of the original starch. A good yield is 6 gallons of absolute aclohol per 100 lb. of starch, as compared with a theoretical yield of 7.16 gallons if the starch could be all transformed into alcohol and carbon dioxide alone. This 6 gallons represents 83.8 per cent. of the theoretical yield, showing therefore a loss of 16.2 per cent. A yield of 6 1/3 gallons, corresponding with a loss of 11.3 per cent., would be a high yield, obtainable under the best conditions; 5§ gallons would represent a medium result, and 5 1/2 would point to the conditions being unsatisfactory somewhere.

In British practice, the average yield of alcohol obtained in ordinary good working with kiln-dried malt and grain may be put at about 65 1/2 gallons of absolute alcohol or 115 gallons of proof spirit per ton of materials.

"Attenuation" yield. - For revenue purposes, in this country a presumptive yield of alcohol is calculated from the quantity of wort, and its specific gravity before and after fermentation. If, for example, 7000 gallons of wort at sp. gr. 1045° (water = 1000) are fermented down until the sp. gr. is 998°, the amount of "attenuation," or " degrees of gravity lost," is 1045 - 998 =47°. It is presumed that for every 5 degrees of gravity lost there is produced alcohol equivalent to 1 per cent, of proof spirit; the quantity of proof spirit thus calculated is the "attenuation charge." Thus in the example taken, the charge is: -

7000 X 47/100 x 5 = 658 proof gallons.

In practice, the yield of alcohol obtained is about 13 to 15 per . cent. more than this. The calculation, however, serves as a rough check upon the quantity of alcohol actually obtained, and in special circumstances the charge of duty may be based upon it.

The tabid given below, calculated from data given by Ullmann,1 shows the average yield said to be obtained in Continental practice:

The following table, showing the yield which may be obtained practically from various materials, is due to T. H. P. Heriot 2: -