New York Agricultural Experiment Station. 187 



same graphs and tables show the moisture content of the samples 

 (referred to dry basis). The atmospheric temperature (also shown 

 ki Table VII) is plotted in the two graphs. The atmospheric tem- 

 perature was obtained by averaging the daily mean temperature 

 for each week. 



Table III. — Bacterial Counts of Potted Soil. 

 Samples taken at Geneva, N. Y., 1912-13. 



The most significant samples taken in 1912-13 are those of Feb- 

 ruary 19th, which were from well-frozen soil. The analyses show a 

 decided increase in germ content over all counts from the soil while 

 unfrozen. The only other samples of frozen soil taken were on 

 February 11th, ten days after the freeze and apparently before the 

 bacteria had begun to increase in numbers. The results, therefore, 

 bear out previous work, but depend upon too few determinations to 

 be conclusive. It is interesting to notice that increases and decreases 

 in numbers of bacteria throughout the experiment accompany rises 

 and falls in the moisture content. 



The results of the work in 1913-14 are more conclusive. During 

 this year eight samples were taken of soil that had been frozen at 

 least two weeks. Of these, all but one gave strikingly higher counts 

 than those obtained in the fall or spring. The results plainly cannot 

 be ascribed to changes in moisture content, for no increase in the 

 latter occurred until the final thaw in March. 



Considering both years' work together, we find that nine out of 

 twelve samples of frozen soil were abnormally high in germ content. 

 Of the three that were no higher than normal, two were taken so 

 soon after the freeze that the bacteria probably had not had time 

 to increase in numbers. The chief significance of this experiment 

 lies in the fact that the possibility of bacteria rising from lower 

 depths during the process of freezing was excluded, thus showing 



