284 BOTANICAL GAZETTE [OCTOBER 
It seems probable that this is accounted for by the inaccuracy of 
weighing and transferring so small a quantity, an error which 
becomes relatively very great in so small an amount. If we 
eliminate this result, it is quite evident from the table that the 
nitrogen content of asparagin as shown by the analyses is a little 
lower than the calculated amount. The difference is very small 
and without much doubt is attributable to slight impurity of the 
salt. Considering the 50 and 100 mg. analyses only, we see a varia- 
tion in the different analyses, not made at the same time, of only 
0.28 of r percent. This seems a good degree of accuracy for such 
determinations. Of course, the per cent of error in quantities as 
small as 3 mg. of the asparagin salt would have little significance. 
Considering the absolute variation of the analyses, including 20 mg. 
and below (not including no. 7), we see only o.1 mg. difference 
between the maximum and minimum variation from calculated 
results. This is true even with slightly impure asparagin. These 
quantities of nitrogen are, it may be observed, larger than those 
dealt with in the nitrogen-free cultures. It would seem safe to 
say then that the limit of error of the method was not far above 
©.Img. in each determination. After considerable experience 
with determinations under different conditions, it is believed that 
this limit might run as high as 0.3 mg. in a long series of determina- 
tions. 
The weighing up of the mycelium in the cultures was done, as 
were, all other quantitative weighings, on an exact Becker analytical 
balance which weighed accurately to o.2mg. The filters were 
first dried in an oven at 100° C. until they showed a constant weight 
(about 6 hours). The culture was then filtered and the filter con- 
taining the mycelium was dried to a constant weight. Glass- 
Stoppered weighing bottles were used for all this weighing and 
drying. As previously stated, the filters were found by a special 
analysis to be nitrogen-free. 
Analyses were also made to test the accuracy of the method 
when nitrates were present. F irst, an analysis was made of 400 Mg- 
of dry ammonium nitrate. This was the amount contained in the 
nitrate cultures of the highest concentration. Next, an analysis 
was made of the original culture solution from which all the cul 
