Aug. 19, 1918 
Reaction on Nitrogen-Assimilating Bacteria 
321 
lowing strains of alfalfa (Medicago saliva) and lupine (Lupinus sp.) were 
stained for flagella: 
Alfalfa 1, 5, 6, 7, 8, peritrichous flagella. 
Lupine 19, single, or rarely two, flagella. 
The shape and general structure of the flagella of the lupine organism 
were different from those of the alfalfa organism. For instance, the 
flagella of lupine 19 are not so long and wavy as those of alfalfa. 
EXPERIMENTAL PROCEDURE 
It was realized from the beginning that the success of this study 
depended largely on the number of parallel tests and the number of dif¬ 
ferent strains of bacteria employed. Therefore each experiment was 
repeated several times. In order to have comparable results all of the 
organisms were grown on the same medium, the composition of which 
is given below: 
Mannitol (CeH^OH)^. 10. o gm. 
Magnesium sulphate (MgS0 4 +7H 2 0). o. 2 gm. 
Dibasic potassium phosphate (K 2 HP 0 4 ). o. 2 gm. 
Sodium chlorid (NaCl). o. 2 gm. 
Calcium sulphate (CaS0 4 +2H 2 0). o. 1 gm. 
Distilled water. 1,000. o cc. 
Only the purest chemicals and conductivity water were used in pre¬ 
paring the culture medium. The reaction of the medium was usually 
neutral to phenolphthalein, although in some of the tests a very small 
amount of alkali was required to make it neutral. After dividing this 
culture solution among a series of flasks, the portions were sterilized and 
adjusted to different reactions with N/io or N/20 acid and alkali. The 
normality of the culture medium is shown in the tabular data. 
In beginning the experiments the acid and alkali limits of growth as 
determined by previous investigators were tried, and repeated tests were 
made until the critical point for the growth of the particular organism 
was reached. 
Because of the importance of the acid-soil problem and the use of 
legumes in an acid system of agriculture, the greater part of this paper 
is concerned with the relation of legume bacteria to acidity, while their 
relation to alkalinity has received only limited study. With the excep¬ 
tion of certain preliminary experiments, consideration was given not only 
to the total quantity of acid added but also to the effect of this acid 
on altering the hydrogen-ion concentration of the medium. Because 
of the low content of buffer substances in the mannitol medium, only 
small quantities of sulphuric acid were required to alter its hydrogen- 
ion concentration. 
To determine the nature and extent of the buffer effect, preliminary 
tests of the hydrogen-ion concentration of the culture medium were made. 
For this purpose, a series of flasks of the medium was prepared in such 
