448 STATE EOARD OF AGRICULTURE. 



mg;. = 4 mg-., which contradicts the fact that only 1.7S mg;. nrea was 

 found to have been decomposed. The new fonunhi gives a fermenting 

 capacity of 66 x 10"^** mg.. only ouetifth of that computed by Burchard, 

 and the amount of urea fermented by the last generation is 0.885 mg. 

 From this, it is evident that the preceding generations fermented 0.442 

 mg.. 0.221 mg.. 0.110 mg., 0.0.5o mg., 0.027 mg.. etc. The sum of these 

 first five data gives 1.74 mg. of urea. 



Exactly the same formula with exactly the same errors has been 

 applied later by Haake to the acid fermentation by B. acidi Idctiei 

 Hueppe (Archiv f. Hygiene Bd. 42, p. 46). Besides the mistake in the 

 formula, both these authors make an error in the arrangement of their 

 experiments, letting the cultures develop for 72 hours at 37°C, Natur- 

 ally, the bacteria are already dying and the maximum number is long 

 passed when the bacteria are counted. This accounts for many rather 

 strange conclusions, especially for the statement that the slower the 

 fermentation, the faster the division of cells. The basis of these con- 

 clusions which can be found in several text books is not sound enough 

 to uphold these views, and a revision will be necessary. 



III. DISCUSSION OF THE ASSUMPTIONS IN THE PREVIOUS CHAPTER. 



In the computation of the above formula, three assumptions have been 

 made which may not be accepted by all bacteriologists. These assump- 

 tions are: (1) That bacteria multiply in geometrical progression; (2) 

 that bacteria cause fermentation as soon as they are transfeired into 

 a fresh medium; (3) that the cell as counted by the plate method in 

 young cultures is the most advantageous unit. These assumptions re- 

 quire a justification. 



The Multiplication of Bacteria in Geometrical Prof/ression has been 

 given considerable attention. It has been found that this process is 

 not wholly reliable for there are too many possible deterrent factors. 

 A number of investigators* have found that immediately after being- 

 transferred, bacteria grow very slowly, but after a few hours, reach a 

 normal rate of multiplication which remains nearly constant for a 

 certain period. This initial retardation is too insigTiiflcaut to have any 

 appreciable influence upon the computation of the fermenting power, 

 since it takes place at a time when the number of cells is too small to 

 form measurable quantities of fennentation products. 



Later Barbert demonstrated by direct microscopic counts of Bacillus 

 coli, that actually there is no such retardation. The findings of the 

 other authors are probably due to an error caused by the plate method. 

 Since, however, all data in this paper are obtained by this method, it 

 is showu in the following pages that this apparent retardation is of no 

 avail. 



*Baseiiau. Archiv. f. Hj-giene Bd. 2.3 (1895) p. 44. 



MuUer. Zeitschr. f. Hygiene, Bd. 20 (1895) p. 245. 



Kossowicz, Zeitschr. f. d. landw. Versiichswesen Oesterreichs Bd. 6 (1903) p. 275. 



Rahn. Centralbl. f. Bakt. II, Bd. 16 (1906) p. 417. 

 tJoumal of Infectious Diseases Vol. 5 (1908) p. 379. 



