QUANTITATIVE MEASURE OF BACTERIAL METABOLISM 79 
liquefied by their action through sterile unglazed porcelain filters. 
Although the bacteria which elaborated the enzymes are removed 
by this filtration, the sterile filtrate still contains the active enzyme 
which will liquefy considerable amounts of gelatin. The function of 
these enzymes is to prepare the gelatin for assimilation by the proteus 
baciilus: the gelatin is broken down by enzyme action to gelatin 
peptone or even to polypeptids. The proteus bacillus does not pro- 
duce a soluble, gelatin-splitting enzyme in gelatin containing utilizable 
carbohydrate, although sugar-free gelatin contains it in considerable 
amount. These gelatinases, however, will liquefy sugar-gelatin quite 
as readily as sugar-free gelatin, indicating that the enzyme itself is not 
inactivated by the sugar, at least in the amount usually employed, 1 
per cent. The same phenomenon is observed in cultures of the cholera 
vibrio and many other bacteria which liquefy sugar-free gelatin 
according to the extensive investigations by Auerbach.^ and of Kendall, 
Day and Walker.- Kendall, Cheetham and Hamilton'^ have published 
experiments indicating that utilizable carbohydrate prevents the 
formation of the gelatin liquefying enzyme. Deamination, a measure 
of the intracellular utilization of the products of liquefaction, is an 
independent phenomenon. These observations indicate how funda- 
mentally the metabolism of bacteria is influenced by the natin-e and 
composition of the substrate upon which they are grown. 
Vn. QUANTITATIVE MEASURE OF BACTERIAL METABOLISM. 
It is possible to measure the nitrogen metabolism of bacteria under 
varying conditions with a very considerable degree of accuracy in 
spite of the minute amounts of products involved. Such measurements 
are not only indicative of the nature and degree of the decomposition 
of purely nitrogenous substances by bacteria; they furnish quantitative 
evidence of the extent of the utilization of carbohydrates by bacteria 
in preference to nitrogenous substances for energy (catabolic) purposes; 
that is to say, such measurements evaluate the nitrogen metabolism of 
bacteria in purely protein solutions, and their metabolism in media 
containing both protein and utilizable carbohydrate. 
Such determinations upon a considerable series of bacteria have 
shown that the bacteria commonly encountered, both pathogenic and 
non-pathogenic, utilize derivatives of protein intermediate between the 
simple amino-acids on the one hand, and the more highly organized 
compounds of the general character of proteoses on the other hand.^ 
All those organisms which ferment glucose produce less ammonia 
in the glucose medium than in the corresponding sugar-free mec'ium, 
although the numbers of living bacteria were found to l)e greater in 
the former than the latter. The relatively small amount of annnonia 
1 Arch. f. Hyg., 1897, 31, .311. -' Jour. Am. Chem. Assn., 1914, 36, 1962. 
3 Jour. Infec. Dis., 1922. 30, 251. 
* For methods see Kendall: Jour. Infec. Dis., 1922, 30, 211, 
