ANNUAL REPORT, 1941 19 



cows and from a full lactation period of continuous feeding of urea to eight other 

 cows in comparison with a similar group continuously fed the regular herd ration. 

 Final conclusions cannot be drawn until the second year of continuous feeding of 

 urea to the eight cows, just referred to, has been completed, and until at least a 

 year's results are available from a group of cows more recently put onto a control 

 ration containing no urea. This last phase of the investigation has been in- 

 cluded in order to check more closely the adequacy of basal protein levels in the 

 ration. All things considered, to date the urea ration seems to be producing 

 results similar to those obtained on the regular ration. 



DEPARTMENT OF BACTERIOLOGY 

 Leon A. Bradley in Charge 



Nitrification in Soils Containing Plant Residues of High Lignin Content. 



(James E. Fuller, cooperating with the Agronomy Department.) During the 

 growing season of 1940 thirteen plots were under observation. During the 

 growing season of 1939 one of these plots had remained fallow, and each of the 

 remaining twelve had been planted with a forage crop. There were no duplicates. 

 Then, in 1940, the whole area was planted with tcbacco, after the plant residues 

 of the preceding crop had been plowed under. Soil samples were taken in the 

 spring of 1941, in mid-season, and again after the harvest. The soils were studied 

 for their ability to nitrify their own nitrogen, added dried blood, and added 

 ammonium sulfate, respectively. The results of the nitrification studies were 

 compared with the quantity and quality of tobacco produced on the plots in 

 1940. There was some evidence, in the dried-blood study, that plots giving less 

 active nitrification gave poorer quantity and quality of tobacco. 



The study was repeated on a second field in 1941. The set-up was replicated 

 six times, giving 78 plots instead of the 13 studied in 1940. Results of 1941 have 

 not been analyzed sufficiently to permit any statement at the present time. 



Comparative Study of Certain Media Employed for Fecal-Flora Studies. 



(James E. Fuller and Irwin Fried.) Much of the investigational work on fecal 

 bacteriology, in connection with nutritional studies, is based upon determining 

 the ratio of bacteria of the lactobacillus group to those of the coliform group, 

 because a predominance of lactobacilli is considered desirable for intestinal 

 health. It is desirable, also, to differentiate the members of the coliform group 

 present in order to make a useful interpretation of results. 



The present study compared certain media commonly employed to enu- 

 merate coliform bacteria. Results were as follows: litmus-lactose agar, brom- 

 cresol-purple agar, and lactose agar with Andrade's indicator were not selective 

 for the coliform group of bacteria, and gave no differentiation within the group. 

 Endo's agar and E. M. B. (eosin-methylene blue) agar gave distinctive colonies of 

 the group and good differentiation, but both produced substantially fewer col- 

 onies than lactose agar when plates were made with these three media from pure 

 cultures of bacteria of the coliform group. This indicates that both Endo's and 

 E. M. B. media give low counts of coliform bacteria when they are used in fecal- 

 flora studies. Certain combinations of non-coliform bacteria produced reactions 

 on these media that could be confused with those of the coliform group. MacCon- 

 key's bile-salt agar inhibited growth of bacteria of the group to a greater extent 

 than did either Endo's or E. M. B. media, and did not give satisfactory differen- 

 tiation within the group. 



