896 Journal of Agricultural Research vol. vi, No. 23 



At the end of the incubation period each soil was transferred with 250 

 c. c. of distilled water to a i-pint Mason fruit jar. Two gm. of powdered 

 lime were added and the jar placed in the shaking machine for 10 minutes, 

 after which it stood in the closed jar until clear. This never required over 

 two hours. At the end of this time an aliquot part, 100 c. c, was meas- 

 ured into a flask and the nitrates determined by the aluminum reduction 

 method (5). 



The nitrogen-fixing powers of the soil were made by weighing 5-gm. 

 portions of the soil into 500 c. c. Erlenmeyer flasks containing 100 c. c. of 

 Ashby solution. These, together with sterile blanks, were incubated for 

 18 days and then the total nitrogen determined by the Kjeldahl method. 

 All determinations were made in triplicate. 



POT EXPERIMENTS 



Dry soil, to a depth of 12 inches, was taken from one of the unmanured 

 plots of the Greenville Farm and very carefully mixed and used as the 

 soil for the pot experiments. This soil, together with the required quan- 

 tity of well-rotted barnyard manure, was packed into the pots. Moisture 

 determinations were made upon the mixtures and then sufficient water 

 added to make up to the required moisture content. The pot and 

 contents were weighed and the moisture content made up weekly to the 

 initial content. The pots were kept on shelves within the building for 

 four months, and then the various determinations were made on each 

 sample as outlined. The temperature of the soil was taken each time 

 before making it up to the moisture content. The manure was applied 

 at the rate of none, 5, 10, 15, 20, and 25 tons to the acre. An acre of soil 

 was considered as weighing 2,000,000 pounds. Bach ton of the manure 

 contained 738 pounds of dry matter, 3.04 pounds of phosphorus, 13.70 

 pounds of potassium, and 16.08 pounds of nitrogen. The moisture was 

 kept at 12.5, 15, 17.5, 20, and 22.5 per cent by weight. Duplicate pots 

 were used in every case with each specific treatment. At the end of the 

 experiment three separate analyses made on each pot, so that each 

 reported result is the average of six closely agreeing determinations. The 

 results are given in Table II. 



The number of bacteria developing on synthetic agar does not seem 

 to have been greatly influenced by the various treatments. All counts 

 are comparatively low. If, however, we average the results for all pots 

 which received the same manurial treatment we find a greater number 

 developed from the soils which received 25 tons of manure to the acre 

 than from any of the others. Moreover, there is an appreciable difference 

 in favor of those soils receiving from 10 to 20 tons per acre over the un- 

 manured soil. The irrigation water apparently depresses the number of 

 organisms, for the greatest number developed from soil receiving the 

 least water; but here also the difference is not marked or regular. The 



