130 The Philippine Journal of Science i9i8 



it is, we can easily see the effect it would have upon the de- 

 terminations in this group alone. The Hiss-Russell bacillus 

 depends upon maltose for its differentiation, and if the maltose 

 is partly inverted this type will not be identified. An investi- 

 gator using a partly inverted maltose would be led astray so far 

 as identifying this type of dysentery bacillus is concerned, while 

 if he tried to determine the reactions of the dysentery bacilli as 

 a class to the three carbohydrates, glucose, mannite, and maltose, 

 he would find his Shiga strains fermenting glucose, not ferment- 

 ing mannite, but fermenting maltose. Mannite tends to preserve 

 its identity, while maltose is unstable and tends to change. 

 The inclination would be to explain this upon the basis of 

 inconstancy in the fermentative action of the bacterium, whereas 

 the cause is to be found in the impurity of the carbohydrate. 

 Long experience has shown that the dysentery bacilli as a class 

 ferment the three substances named above in the following order : 

 First most readily glucose, next mannite, and third and least 

 readily maltose. Ohno,(3) in a study of types of dysentery bacilli 

 in 1906, found a number of organisms which, in acting upon 

 these three cai'bohydrates, skipped mannite, but fermented mal- 

 tose. He makes no mention of having ascertained whether or 

 not his carbohydrates were pure. If he had his maltose partly 

 inverted to dextrose, it would explain why many of the sera 

 produced by his acid types agglutinated nonacid types just as 

 well as they did the organisms used in their production. 



Castellani(4) in a recent article publishes a table of cultural 

 characters for certain intestinal bacteria, in which he states 

 that the Shiga bacillus may or may not ferment maltose. In 

 the same table the Hiss-Russell bacillus is put down as producing 

 acid in maltose. Now the nonfermentation of maltose is the 

 one differentiating character of the Hiss-Russell bacillus, and 

 if we remove that it becomes a Flexner type. I have found in 

 the series of organisms with which I have worked that several 

 which were at first thought to be Flexner strains were really 

 Hiss-Russell strains. They could be made to give the fermen- 

 tations of the Flexner bacillus by simply changing from Merck's 

 to Kahlbaum's maltose. It is believed that this instability of 

 maltose has been generally encountered and that many of the 

 varying results obtained by different observers may be attributed 

 to it. If we recognize this same principle as applying also to 

 other carbohydrates in different degrees, we can easily see that 

 the promiscuous use of carbohydrates will, when no control ia 

 used to detect impurity or inversion, lead to wide differences in 



