62 HATCH EXPERIMENT STATION. [Jan. 



lent, a decrease of 31 points in the Reichert-Meissl number, 

 and an increase of 9 points in tlie iodine number, while the 

 melting point of the fat remained unchanged. 



9. An anal3^sis of the butter fat will seldom give positive 

 knowledge concerning the firmness or body of the butter. 



10. A high iodine number is indicative of a soft or " light- 

 bodied " butter ; but a high melting point is not a sure indi- 

 cation of a hard, firm butter. 



11. It seems probable that neither the proteid or car- 

 bohydrate groups, when fed in normal ainounts, have any 

 noticeable influence in changing the proportions of the sev- 

 eral milk ingredients, or in modifying the chemical charac- 

 ter of the butter fat ; such changes, so far as they occur, are 

 due to the presence of the oil in the feed stuff. 



12. Corn gluten meal with a minimum percentage of oil 

 produced rather a soft, yielding butter ; this condition was 

 noticeably increased by the addition of corn oil to the ration. 



13. The flavor of butter depends primarily on the cleanli- 

 ness of the milk, stage of lactation of the animal, method of 

 butter manufacture, and especially upon the character of the 

 starter employed. Normal feeding stufls are of secondary 

 importance in establishing butter flavor.^ 



14. The present and previous experiments indicate that 

 starchy feeds produce a hard-bodied butter, vegetable oils a 

 soft butter ; some proteids a hard-bodied butter, others but- 

 ter of a softer, lighter body. 



1 A possible exception to the above may be made in the case of young, clean 

 pasture grass. 



