ATMOSPHERIC-XITROGEN FERTILIZERS. 
phate rock for the purpose of rendering cyanamid more generally- 
suitable for use in mixed fertilizers. The material, which has been 
made only on an experimental scale, contains about 5.8 per cent 
nitrogen, approximately 65 per cent of which is present as urea and 
ammonium sulphate, 30 per cent as guanylurea sulphate, and 5 per 
cent in undetermined forms. Nearly one-half of the phosphorus is 
insoluble in standard ammonium-citrate solution. 
Calcined phosphate is the product obtained by heating a mixture 
of phosphate rock, carbonaceous material, and an alkali salt, such as 
salt cake or niter cake, in a rotary kiln. A high percentage of the 
phosphorus can be obtained in ammonium-citrate soluble form in 
this manner. Unlike acid phosphate, this material can be mixed 
with cyanamid in any proportion without producing deleterious 
changes. The new material has not as yet come on the market. 
SOILS AND CROPS. 
In choosing the areas to be used for experimental purposes the 
necessity of limiting selections to the two Government reservations 
at Muscle Shoals, Ala., comprising about 1,000 acres, proved a 
handicap. This soil is for the most part poor and is rather hilly and 
irregular. The most suitable areas from the standpoint of topog- 
raphy and also uniformity were naturally the most fertile and less 
adapted to fertilizer experiments. This irregularity naturally 
minimized the value of the results to a certain extent, but nevertheless 
the responses to fertilizer applications were usually sufficiently 
striking to leave, no doubt as to their value. In this connection it is 
important to emphasize the value of the observations made during 
growth. The effects of a given treatment on germination and early 
and late growth were nearly always quite striking regardless of 
natural fertility and gave a better picture in many instances of the 
real merits of the material than did the yields. 
The total area used for experimental purposes was approximately 
10 acres the first year and 20 acres the following two years. All the 
work was done on areas of about 5 acres each. Three soil types, 
namely, Clarksville, Colbert, and Decatur loams, were represented. 
These were not widely different either in chemical or physical nature, 
all being poor in nitrogen and phosphorus except in the lower areas, 
which had benefited by leachings from higher levels. 
The lay-out of the plats on the various experimental fields is shown 
in the diagrams (see figs. 1 to 4). The size of the plats varied from 
one-fortieth to one-twentieth of an acre. 
A variety of crops was grown the first year, primarily to determine 
the responses from treatments of cyanamid and ammonium nitrate. 
These included several vegetables, tobacco, grasses, small grains, 
forage crops, cotton, and corn. The experiments of the last two 
seasons were confined to cotton and corn, the crops which, together 
with tobacco, receive the bulk of the fertilizer used in the South. 
METHODS USED IN EXPERIMENTS. 
The fertilizer mixtures used were of such a variable nature that 
reference must be made to the tables given later to determine the 
particular combinations tested. In these tables the term fertilizer 
ratio refers to pounds of N, P 2 5 , and K 2 in the 1919 results and to 
