66 



MASS. EXPERIMENT STATION BULLETIN 158. 



The difference between the American and German figures is within the 

 limits of variations in different samples.' 



(2) Digestion Coefficients with Sheep. 



It will be seen that the product used by Kellner for some reason was 

 much less digestible than that used by ourselves. 



Pounds Digestible in 100 Pounds of Shells. 



Protein, 



Fiber, 



Nitrogen-free extract. 



Fat 



Total, . 



31.45 



It is quite e\'ident that the proteid matter is only slightly digestible 

 and may be considered a negligible quantity ; hence the value of the prod- 

 uct consists of the digestible fiber, fat and extract matter. On the basis 

 of his results, Kellner remarks that the cocoa shells have no more feed- 

 ing value than straw. 



The net available energy on the basis of our own digestion trials is 63 

 as compared with corn meal equal to 100. When, however, one considers 

 their non-palatability and their rather objectionable appearance, to- 

 gether with the results of other investigations, it does not seem advisable 

 to rate them as having more than one-half the feeding value of corn meal. 



(3) Feeding Trials. 

 A number of milch cows were fed from 2 to 3 pounds, daily, of the cocoa 

 shells, both ground and unground, mixed with other grains. One cow 

 was induced to eat as high as 5 pounds when mixed with malt sprouts and 



1 Foreign workers have shown the presence of the alkaloids caffein and theobromine, also a 

 considerable percentage of pentosans. Fowler, in the laboratory' of the Massachusetts Agricul- 

 tural College, has determined the percentages of the alkaloids, and has found also 8.3 per cent, 

 of pentosans, 7.3 percent, of galactans, a little over 1 per cent, of starch and traces of sugar. 



' Obtained by Kellner. 



