394 L. D. BUSHNELL 



Since there were such large amounts of ammonia present, and 

 because it is so difficult to remove it from the sample, corrections 

 by the use of the factors have probably given results as nearly 

 correct as could have been obtained by any other method. 



Volatile fatty-acid determinations. For the determination of 

 volatile fattj^-acids we have followed the Duclaux method as 

 modified and described by Dyer (1916). This author suggested 

 that the method might be employed in the studj' of acid produc- 

 tion by bacteria, and Wolf and his associates used it for the study 

 of the acid production by certain pathogenic anaerobes. 



While the results which we have obtained show that there is 

 a difference in the fermentative activities of this organism upon 

 the different carbohydrates the differences are not great. The 

 distillation of volatile fatty acids is not an easy task and while 

 its use may be of value in detecting in a general way what prod- 

 ucts are formed during a given fermentation, their value as a 

 diagnostic biochemical test for bacterial differentiation is limited. 



Our attitude toward this test cannot be better expressed than 

 by quoting from the summary of Wolf and Telfer. 



An experimental critique of the Dyer method of estimating volatile 

 fatty acids has been made. This method, while perfectly satisfactory 

 in the form stated by its author in dealing with a mixtm'e of two vola- 

 tile acids, the nature of which is Icnown, fails when a mixtm-e of unknown 

 acids is to be analyzed. The color tests as proposed by him are satis- 

 factory when dealing with pure acids, but not as positive as could be 

 desired for the identification of an acid in the mixture. The separation 

 of the acids is necessary before any reliance can be placed on these 

 color tests. 



From a study of table 4, it may be seen that in most cases the 

 fractionation curves lie very near to those of propionic and buty- 

 ric acids. By the tests devised by Dyer we could obtain no evi- 

 dence of propionic acid, though various fractions were refraction- 

 ated as indicated by Wolf and Telfer. We could obtain some 

 fairly good tests for acetic and butyric acids after repeated rc- 

 fractionations; in the case of milk, the odor would indicate vale- 

 rianic acid though no quaUtative test could be obtained for this 

 or for caproic acid. 



