580 Journal of Agricultural Research vo1.xv.no. n 



It has likewise been demonstrated from the study of many classes of 

 silage of high quality that the general course of development of the total 

 microbial flora is slow to rapid, usually covering a period of a few days to 

 two weeks, followed by a gradual decline. The data in Table I in- 

 dicate a like tendency, but the characteristic development is not so pro- 

 nounced as would be expected from a good grade of silage. This is par- 

 tially explained, however, by the method used in sampling and also by 

 the small capacity of the silos used, which did not favor optimum silage 

 fermentation. The method of sampling, no doubt, accounts for a large 

 portion of the variations noted, since the method did not always allow 

 the collecting of proportionate amounts of leaves and stem, or where 

 supplements were added of proportionate amounts of alfalfa and the 

 supplement. 



BACTERIOLOGICAL INVESTIGATIONS IN 1915 



The seven silos were again filled in the spring of 191 5 with alfalfa and 

 the supplements previously mentioned. The general plan of study was 

 the same as for 191 4. 



However, glucose-litmus broth was used to determine the total number 

 of acid producers. It was prepared by using i per cent glucose broth to 

 which a few drops of litmus solution had been added. Several tubes of 

 this medium were inoculated with different dilutions of the silage in- 

 fusion and incubated. The total acid producers were determined by 

 noting the acid reaction in the highest dilution present. 



This method was used in place of a litmus-agar medium because a 

 more accurate estimate of the acid producers could be obtained. If on a 

 litmus-agar plate the acid colonies are few and other types predominate, 

 the colonies may fail to appear acid on account of the neutralization of 

 the acid by the alkaline by-products from the other types. The or- 

 ganisms in the glucose-broth solution are not held in one place, as they 

 are in the solid-agar media. This, together with the fact that the glu- 

 cose broth acts as an enrichment medium for the acid group, gives better 

 opportunity for the acid bacteria to increase more rapidly than the mis- 

 cellaneous organisms. Likewise, the acid produced from the more rapid- 

 growing acid bacteria is sufficient to check the slower development of the 

 miscellaneous types. The dilution method is more tedious, but, pro- 

 vided the differences between the dilutions are small, the results ob- 

 tained are more satisfactory and more accurate. Plain gelatin was used 

 to determine the protein-digesting types. Not having a satisfactory 

 place for the incubation of gelatin plates, sterile tubes of gelatin were 

 inoculated with varying dilutions of silage, and incubated at 37.5° C. for 

 five days. The number of liquefiers were determined by placing the 

 gelatin tubes in the ice box, after their removal from the incubator. 

 From previous experience it has been found that a tube of digested gela- 

 tin will not solidify on cooling. Hence, by noting the highest dilution 

 showing liquefaction after cooling, the number of liquefiers can be esti- 



