288 Journal of Agricultural Research voi. x, no. 6 



and rye, 2 to i ; alfalfa alone; and rye alone. Alfalfa and corn, and alfalfa 

 and molasses made satisfactory silage for dairy cows. The sweet-sor- 

 ghum butts as a supplement would no doubt have made excellent silage 

 if they had been first class. The rye proved unsatisfactory as a supple- 

 ment. 



In the center of each silo was placed a telescopic can holding from 25 to 

 50 pounds of material. The construction of these cans was such that the 

 silage in the cans was subjected to the same pressure as the rest of the 

 silage.* On opening it was found that the silage in each can was better 

 than the silage next to the can in the silo. This proves that rigid exclu- 

 sion of air is of prime necessity, but such rigid exclusion of air is not pos- 

 sible under practical conditions. 



CHEJMICAI, ANALYSES 



Samples for chemical analyses were taken by boring into the side of 

 the silo with a 2 -inch auger and then removing a core of silage with a 

 i^-inch auger. The boring extended to about the center of the silo, 

 and gave a sufficiently large sample for chemical work. This method of 

 sampling is not satisfactory, though it is probably the best that can be 

 used under such circumstances. This is especially true when the silage 

 is made of a mixture of alfalfa and some supplement. Then, when the 

 silage is partly spoiled in places, there is no certain way to ascertain 

 whether an undue proportion of the bad silage is included in the sample. 

 In addition to this, the leaves and stems differ very much in composition, 

 and the right proportion of these is not always obtained. This difficulty 

 in obtaining a satisfactory chemical sample should be considered when 

 the chemical data are studied. But for this difficulty in sampling, the 

 results would show more uniformity in variation. 



The plan was to take samples from each silo every day after filling for 

 the first seven days, every other day for the next week, then every four 

 days for the next two weeks, then once a week, and finally once a month. 

 The samples were prepared for analysis by passing them through a clover 

 cutter whose feed was so regulated as to cut lengths H inch and less. 

 This reduced the material to a satisfactory degree of fineness for all 

 analyses made of the silage in the condition in which it came from the 

 silo. The total nitrogen, as well as the complete feed analysis, was deter- 

 mined on the dried, finely ground samples. The following determina- 

 tions were made on the fresh samples: Moisture, total acidity, sugar, and 

 nitrogen in amino form as determined by the formol-titration method. 

 These results are tabulated in Table III. 



' These cans were designed by Prof. J. T. Willard about 25 years ago, while making silage experiments 

 at that time. 



