594 



Journal of Agricultural Research 



Vol. XII, No. 9 



the outside air. However, the more rapid increase during the early 

 part of the period was entirely independent of this factor. A compari- 

 son of the temperature records secured in this experiment with the data 

 which have been obtained on ordinary silage at this and other Experi- 

 ment Stations indicates that there is no wide difference, if any, between 

 the rate and amount of increase in temperature in silos in which ordi- 

 nary silage and corn-stover silage have been stored. 



BactERIOIvOGICAIv observations. — A quantitative bacterial exami- 

 nation was made on each of the samples taken. The juice obtained was 

 plated in proper dilutions and the counts obtained were reported as 

 numbers per cubic centimeter of juice. Lactose agar was used, and the 

 plates were counted after six days' incubation at 33° C. As may be 

 seen from Table III, it appears that the bacterial comit increases during 

 the first week and is followed bv a continued decrease thereafter. 



Table III. — N timber of bacteria in stover silage at different stages of curing 



Direct microscopic examinations of the silage juice were made in 

 order to follow in a general way any marked changes which take place 

 in the bacterial flora during the curing process. This at best could only 

 give suggestive data, but such examinations are sometimes important 

 in connection with cultural studies. At first a great variety of cells 

 were observed. During the first two weeks rods and cocci were appar- 

 ently present in about equal numbers, after which the rods became 

 increasingly predominant. Toward the end of the experiment prac- 

 tically nothing but rods were found in the microscopic preparations. 

 Because of the high acidity it is not likely that the cocci were active 

 nearly as long as they appeared under the microscope. But it is prob- 

 able that the acid medium would tend to peserve the cells so that they 

 would appear for some time after they were inactive or even dead. 



A qualitative bacterial study was also carried out. All of the colonies 

 from a representative lactose-agar plate from each sample were isolated 

 and subjected to a cultural study. For the present purpose they may be 

 divided roughly, according to their action on litmus milk, into acid- 

 forming, casein-digesting, alkali-forming, and inert groups. The acid 

 formers may be further divided according to whether they produced 

 sufficient acid to cause coagulation of the milk. The distribution of 

 these groups in silage at various stages of its fermentation is shown in 

 Table IV. 



