124 Reporr or DEPARTMENT OF BACTERIOLOGY OF THE 
plained (page 106), there were 28 errors out, of 158 trials, or 
18 per ct. This decrease of 15 per ct., or practically one-half, 
in the inaccuracy of the test brings out clearly the importance 
of revivifying cultures before attempting to determine their 
fermentative ability. 
Again it will be noted that the failure to form gas during 
a part of the tests is not equally distributed among the three 
sugars, but. that there are 47 cases with dextrose, 17 with 
lactose and 27 with saccharose. While it is true that a few 
more tests were made with dextrose than with either of the 
other sugars, this increase is not at all proportional to the 
larger number of failures. 
In arranging Table III the organisms which had always 
formed gas from all of the sugars at all of the trials were 
placed first, followed by those in which failures to form gas 
had been noted. These latter were arranged in the order of 
the frequency of this failure down to Miller: Stalk 3 No. 1 
which, while it formed gas from each of the sugars during 
some of the tests, failed to do so in 18 out of 21 instances. 
The fermentative vigor of these cultures was fairly pro- 
portional to this arrangement. The fairly distinct amounts 
of gas formed by the first cultures in the table gradually di- 
minishes down the table until there is rarely more than a 
small bubble with dextrose in the case of CII, CIII, XCVIII, 
XCIV and Miller Stalk 3 Nos. 1, 2 and 3. The amounts of 
gas formed from saccharose also undergoes a like shading 
down, but its formation in appreciable quantities continues 
beyond the point where the formation of gas from dextrose 
ceases. Beginning with Potter’s Bacillus there is a group 
which does not form visible gas from either dextrose or saccha- 
rose, but is able to form it from lactose. The remaining cul- 
tures in this table have never produced visible gas in the 
fermentation tubes in any of the determinations which have 
been made. 
Riverhead Stalk 3 No. 1 is a significant member of the 
eroup connecting this last group with the one preceding it 
since in the earlier studies of this organism it fermented 
