POSSIBLE SOURCE OF ERROR. 237 



for instance, by the proportion of nitrogen in each, 

 merely, without taking into consideration other prop- 

 erties. In other words, roots may be compared with 

 each other on that basis merely, and grasses with each 

 other, and leguminous plants with each other, but not 

 root crops and grasses. This fact is alluded to as a 

 possible source of error in some of the earlier researches 

 of Boussingault, and not as materially affecting the prac- 

 tical value of the table. 



The mode of using Table XII. is very simple. Good 

 upland meadow hay or what would be called in New 

 England good English hay is taken as a standard of 

 comparison. Now, if we wished to produce the same 

 results with carrots as with one hundred pounds of 

 good, average English hay, we must use, according to 

 Boussingault's column of equivalents, 382 pounds of 

 carrots, or for each pound of hay 3.82 pounds of car- 

 rots ; and, according to the practical experiments men- 

 tioned, 366 pounds, 250 pounds, 225 pounds, 300 pounds, 

 and so on, to each 100 pounds of hay. 



According to the theoretical values of Boussingault, 

 100 pounds of hay are equal in feeding qualities to 65 

 pounds of barley, 60 pounds of oats, 58 pounds of rye, 

 or 55 pounds of wheat. While, according to the exper- 

 iments of Thaer, 100 pounds of hay produced the same 

 effect as 76 pounds of barley, 86 pounds of oats, 71 

 pounds of rye, 64 pounds of wheat. 



With regard to the analyses of Tables VI., VII., VIII., 

 and IX., some slight allowance should perhaps be made 

 for difference of climate, since it is well known that 

 grasses, as well as other plants, grown rapidly in a hot 

 sun, which we usually have in the months of May, June, 

 and July, contain a much larger amount of nutritive and 

 saccharine matter than those grown slower, and in a 

 greater amount of available moisture both in the atmos- 



