300 The Fixation of Nitrogen in Faeces 



tube is, of course, withdrawn when the sample is safely at the bottom 

 of the flask. The following figures are typical of the degree of accuracy 

 possible with this method of sampling : 



Accuracy of Sampling. 



Weight of horse faeces taken (grams) . 

 Weight of nitrogen found (grams) 

 Per cent, nitrogen on fresh sample 



All estimations were made in triplicate and the mean result used in 

 calculating the weight of nitrogen taken for each experiment. 



In experiments of this type it is essential to know what differences 

 are really significant. It was found that with the quantities usually 

 taken, the total error of experiment amounted to about 2-5 % of the 

 original total nitrogen. In the actual experiments no change in the 

 nitrogen content under 5 % can be regarded as significant. 



In the first experiment with horse droppings 20 c.c. of tap-water 

 were added to some of the units of the series, in order to make the liquid 

 surface, exposed to the air currents, correspond with that in some parallel 

 experiments with urine. After 74 days at laboratory temperature 

 (13° C.) no change of nitrogen content had occurred in 7 out of the 9 

 units, but in the remaining two, to both of which tap-water had been 

 added, an increase of total nitrogen of 7 % was recorded. In this 

 experiment no ammonia was volatilized and recovered in the acid traps. 

 The small amounts of ammonia sometimes recovered in later experi- 

 ments are included in the total nitrogen figures in the tables. Although 

 the possibility of nitrogen fixation was recognised, these results were 

 attributed to some experimental error, such as leakage of unwashed air 

 into the apparatus, and no particular notice was taken of the discre- 

 pancies, beyond taking steps to avoid leakage in subsequent experi- 

 ments. In any case fresh horse faeces did not seem at all a favourable 

 medium for the growth of the better known nitrogen fixing organisms. 

 Remy^ has stated that farmyard manure is an unsuitable medium for 

 Azotohacter. He found a loss of nitrogen in his experiments. The 

 Azotobacter group are said to dislike a nitrogenous environment and, 

 if they could have developed freely, would have fixed more nitrogen 

 than the amounts actually found. On the other hand, the Clostridia 

 are chiefly, if not exclusively, anaerobic and the conditions of the 

 experiment were strongly aerobic. Nevertheless the possibility is not 



1 Cenir.f. Bakt. 2 Abt. Bd. xxii, 1909, p. 588. 



