92 UNIVERSITY OF WISCONSIN STUDIES 
nearly identical characters. The other two included only a 
few cultures and were of doubtful validity. 
The value of divisions based on the liquefaction of gelatin 
may be very properly questioned. It is well known that the 
ability to exerete a proteolytic enzym may be lost and Burton 
and Rettger think that there may be an error in determina- 
tions which do not permit an incubation period longer than 
20 days. If the fact of gelatin liquefaction can be established, 
there is no doubt that it indicates a distinct evolution in 
methods of obtaining food supply and consequently has tax- 
onomie significance. The indications are that the liquefiers 
of our high ratio group have other characters in common 
which separate them from the non-liquefiers and we have ac- 
cordingly considered them separately. 
Gelatin - SO - 15-3.0 “and P+ 
Aaonite + Adonite — 
700_80_ ov 40 20 +|- 20 40 60 80 100 100 80 _G)I_40 20 +|- 20 40 60 80 100% 
Indo! 
Saccharose, 
Roffinose. 
Starch 
tnulin 
Mannitol. 
Glycerol. 
Salicin. 
Dulcitol. 
Fig. 4 
While there is a considerable range in the gas ratio of the 
high ratio group, the variation under different conditions and 
between cultures otherwise identical is so great that this 
character cannot be safely used for further subdivision. Many 
groupings on the basis of substances fermented are possible. 
The correlation of adonitol fermentation with a definite source 
has suggested the use of this character as a possible indicator 
of a line of demarcation. 
The differences in the two groups formed in this way are 
shown by Fig. 4, which is based on data given in Table IV. 
The adonite + cultures are more active in every way than 
the adonite — cultures, and are especially so in fermenting 
