Bacteriological Studies of the Fixation of Nitrogen. 33 
well known Black Alkali of the Southwest in which sodium carbonate 
is the active principle in bringing the soil humus into solution, which 
solution bemg highly colored, may give the surface a dark appearance. 
Again there are some soils which contain sufficient quantities of cal¬ 
cium chloride to absorb enough moisture to impart a dark color to the 
soil. I have Dr. Headden’s statement that none of the soils which are 
concerned in this project contain enough of either sodium carbonate 
or calcium chloride to account for this phenomenon. Our problem 
is manifestly different from either of these. 
In our pure culture studies of the Azotobacter flora of these soils, 
we have isolated what appear, in the final analysis, to be six or seven 
varieties of Azotobacter chroococcum. Three of these, at one time 
or another, have produced the characteristic brown color on mannite 
agar. . One of these, No. 3, has maintained this character undiminished 
since it was first isolated; the second, No. 93, acquired the color three 
weeks after isolation, retained it for six weeks and then lost it; 
the third, No. 1, has produced a small amount of a 
light brown pigment at times ever since its isolation but there has 
been nothing constant in this respect until the last six weeks when it 
has begun to produce a heavy dark brown color. Three of the remain¬ 
ing four cultures, Nos. 4, 8 and 10, have been characterized by their 
spreading nature and their abundant, moist, raised, gelatinous, starchy 
white to yellowish growth on mannite agar. Morphologically and cul¬ 
turally, these three possess such differences as seem to make them 
distinct from one another; the fourth, No. 13, differed from all the 
rest in the production of a delicate cream colored pigment; it was 
spreading in habit but flat, not gelatinous and grew only moderately 
well on mannite agar. Unfortunately, this culture was lost early in 
our work, and consequently it has been given no consideration in this 
treatise beyond mere mention. 
The brown color of cultures 1, 3, and 93 served to identify them 
beyond reasonable doubt as Azotobacter chroococcum. The other four 
cultures were left unclassified, for the time being, except to place 
t em in the genus Azotobacter after they had been shown to possess 
nitrogen fixing powers in pure culture. 
The close resemblance between the brown pigment formed by 
some of our cultures and the brown color present on certain soils was 
suggestive to say the least. It seemed reasonable to me that there 
might, be something peculiar to our soils which could stimulate and 
intensify the pigment producing power of Azotobacter chroococcum. 
To. determine this, a number of synthetic agars were prepared, the 
composition of. which was based upon the water soluble salts present 
in a certain nitre soil.. The carbon was supplied in the form of 
mannite. Each agar differed from every other in the omission of one 
°. ^ 1e -rompounds, our object being to determine by elimination, if pos¬ 
sible, if any one constituent was directly responsible for the brown pig¬ 
ment. The analysis of the water soluble salts in the soil which was 
