No. 4.] GERMAN EXPERIMENT STATIONS. 381 



compare the value of all feed stuffs to hay. This was 

 accomplished by veiy simple chemical analysis, in that the 

 portion soluble in water, dilute acid and alkali was held to 

 be the amount valuable as food for the animal. The results 

 ol^tained by this method were arranged in tables. An im- 

 provement were the tables compiled l)y Boussingault, in 

 which fodders were compared by the amount of nitrogen 

 they contained. Wolff went a ste}) further, in that he 

 divided the fodder components into two great groups, — 

 the nitroo-en-containino; substances, and the nitro2"en-free 

 matter. So far had our knowledge progressed when the 

 G()ttino;en station beoan its work. After makins; some ex- 

 periments disproving the " hay theory," Henneberg adopted 

 a plan which has virtually been followed at the Gottingen 

 station ever since ; namely, ( I) resolve the fodders into their 

 chemical constituents, as Liebig had already attempted, and 

 (2) follow these different groups of constituents through 

 the changes they undergo in the animal body, in a similar 

 way as Bischoff and Voit were doing with carnivorous 

 animals at Munich. 



1. Chemical MetJiocU. (Weender MetJiod.) 



Liebig recognized three groups of nutrients in fodders : 

 viz., albuminoids (nitrogen-containing substances), Hits, and 

 carbohydrates (sugars, starch, gums, etc.) ; and later a 

 fourth group was admitted, namely, the cellulose. The 

 methods for determinino; these substances were at first 

 crude, and when it was attempted to taliulate results and 

 secure averages, it was found that the percentages differed 

 so much that they were of very questionable value. Hen- 

 neberg, aided by his assistants, made a critical study of 

 these methods, made many improvements, and finally estab- 

 lished the so-called AVeender method, which, with very 

 slight changes, Is still employed in all the agricultural 

 laboratories of the world. He divided the fodders into 

 four great groups of constituents, and called them crude 

 cellulose, crude fat, crude protein and nitrogen-free extract 

 matter. It was known at that time that the method was 

 by no means perfect, and our information concerning plant 



