12 
AMERICAN JOURNAL OF BOTANY 
[Vol. 8 
The Forster modification, employed in analyzing the media containing 
nitrates, consists essentially in the addition of sodium thiosulphate to the 
digestion mixture. The procedure was as follows: A considerable part 
of the water of the medium was boiled off in the presence of lo cc. of con- 
centrated sulphuric acid, the process being continued until active foaming 
began. When cool, 20 cc. of the phenol-sulphuric acid digestion mixture 
(icq gr. phenol in 924 cc. concentrated sulphuric acid) were added and 
allowed to stand several hours or over night, the mouths of the flasks being 
covered during this time. Two grams of sodium thiosulphate were then 
added, and, after the reaction was completed, i gram of mercuric oxide, 
I gram of zinc dust, 10 cc. of concentrated sulphuric acid, and finally 10 
grams of potassium sulphate were added in the order named. The flasks 
were heated over a low flame until foaming ceased, when the flame was 
gradually increased to full intensity. Digestion was continued for about 
20 minutes after the liquid became clear, the process requiring about 2| 
or 3 hours. The distillation was carried out as with the Gunning method, 
except that before making the solution alkaline the mercury was precipitated 
out by the addition of 25 cc. potassium sulphide solution (20 gr. KoS in 
500 cc. H2O). 
Numerous "blank" determinations of the nitrogen content of the re- 
agents used in the two methods were made, the average of these being 
deducted from the determinations of the cultures and checks. Care was 
taken to use uniform reagents for the analyses of each series. 
In the preliminary determinations no difficulty was experienced with 
the .Gunning method in securing complete recovery of the nitrogen from 
solutions of urea or ammonium sulphate. In the presence of agar or of 
agar and glucose the usual trouble with foaming was encountered, but the 
determinations checked readily within the limits of experimental error. 
The recovery of total nitrogen, including nitrates in the presence of 
agar, was more difficult, but consistent results were obtained with the 
Forster method as outlined above. It was found necessary to allow the 
digestion mixture to stand a considerable length of time (usually over night) 
in contact with the dissolved and partially concentrated agar medium before 
the addition of the other reagents. Following the experience of Duggar and 
Davis (1916), it was also found advisable to permit the flasks to stand about 
an hour after the addition of the sodium thiosulphate. The low results 
obtained when the digestions are completed more quickly may possibly 
be due to the rather extensive dilution of the reagents by the water of the 
culture medium and the consequent slower reduction of the nitrates. 
Tables A and B show some of the rCvSults obtained in the preliminary 
determinations. 
