New York Agricultural Experiment Station. 185 



from below. Both aerated and unaerated soil were used in these 

 pots, the unaerated soil having been obtained by digging a block 

 from the field and transferring it directly to a pot. It was hoped 

 by this means to see whether the failure of the previous pot experi- 

 ment to show an appreciable increase in germ content could have 

 been due to the unnatural aerated condition of the soil. 



Parallel to these pot experiments, tests were made of the same 

 soil (Dunkirk silty clay loam) in the field. In these tests, also, 

 both aerated and unaerated soils were used. The unaerated soil was 

 merely an undisturbed field plat that had been kept fallow since 

 1911, the upper two or three inches having been cultivated after 

 every rain to preserve a dust mulch. There were two portions of 

 aerated soil, one aerated in November, 1911, the other in November, 

 1913. The former portion, after aeration, was replaced within a 

 large tile such as is used for sewer pipe, two feet in diameter and two 

 feet long, buried in the field with only its flange above ground, and 

 the soil so placed within it that the subsoil was below and the surface 

 soil above in a layer of the same depth as occurs naturally in the 

 field. This soil, like that of the unaerated plat, had been kept 

 fallow since 1911. The second portion, aerated in 1913, consisted 

 of surface soil alone and was replaced within a smaller cylinder, six 

 inches deep and six inches in diameter, likewise sunk in the field. 



A third series of tests was made in an entirely different soil, Dun- 

 kirk fine sand. Two spots were chosen, about fifty feet apart. One 

 was in sod; the other at the edge of a strawberry field, in a spot that 

 was practically free from any vegetation, either because it had been 

 frequently cultivated or because of the poor quality of the soil. 

 These tests were made to see whether the bacteria increase in num- 

 bers in a frozen sand as well as in a clay loam. The samples exam- 

 ined were unfortunately few in number. 



METHODS. 



The methods employed have been kept as nearly as possible the 

 same as those used in the earlier work. Soil samples were regularly 

 taken by boring to about six inches, although some of the winter 

 samples were taken to a slightly less depth because of the difficulty 

 in boring through frozen soil. The soil thus obtained was thoroughly 

 mixed, by sifting, if dry enough, or by stirring, if muddy. A 0.5- 

 gram portion of this sample was finally selected and shaken for two 

 minutes with 100 cubic centimeters of sterile water in a stoppered 

 flask. After this shaking, the suspension was further diluted, care 

 being taken to keep the contents of the flask in motion when any of 

 the suspension was withdrawn. One cubic centimeter of the proper 

 dilution was finally added to each plate. The dilutions used were 

 1 : 100,000 and 1 : 200,000 or 1 : 200,000 and 1 : 500,000. Three or 

 four plates of each dilution were always made. 



