170 EXPERIMENT STATION EECOED. 



The increase in both the absolute and relative amount of fat was greater with 

 cows giving a high yield of milk. The chemical constants of the fat were 

 changed, the saponification number being raised while the iodin and solidify- 

 ing point were lowered. The effect of the palm-nut cake on the character of the 

 milk fat was slow, but remained for some time after the cake was discontinued. 

 About 2 kg. of the cake per 500 kg. of live weight of the animal is recommended 

 as a ration for milch cows. 



The effect of feeding with coconut cake and linseed cake on the composi- 

 tion of butter fat, H. T. Cranfield (Analyst, 36 {1911), No. ^26, pp. U5-U9, 

 figs. 2). — Two lots of 4 crossbred Shorthorn cows each were fed a basal ration 

 of mangels, oat straw, chaff, bran, dried grains, rice meal, and hay. As a sup- 

 plement all the cows received during the first week 5 lbs. of undecorticated 

 cotton cake daily, and in the second and third weeks lot A received 5 lbs. of 

 coconut cake and lot B 41 lbs. of linseed cake. These two supplements were 

 reversed the fourth and fifth weeks, and for the last week all the cows were 

 fed as during the first week. 



The average constants of the milk fat during the coconut-cake period were 

 as follows: Lot A, refractive index at 35° C. 44.2, Relchert-Meissl number 29.5, 

 Polenske number 3.3 ; lot B, refractive index 45.4, Reichert-Meissl number 30.2, 

 Polenske number 2.9. During the linseed cake period lot A had a refractive 

 index of 46.2, Reichert-Meissl number 29.1, Polenske number 2.3 ; lot B, refrac- 

 tive index 47.7, Reichert-Meissl number 29.1, Polenske number 1.7. In the 

 cotton-cake period the refractive index was 46.4, Reichert-Meissl number 30.8, 

 and Polenske number 2.3. 



The cakes appeared to have a different effect on the keeping properties of 

 the butter fat. The samples were kept in small corked bottles or flasks, and it 

 was noticed that after several weeks butter made during the linseed and cotton 

 cake periods had turned white and rancid, while that from the coconut cake 

 period showed only a slight change. This preservative effect is an advantage 

 in favor of the coconut cake. There was no very great difference shown in 

 texture and flavor of the two sets of fresh butter, although as a rule the butter 

 from the coconut cake was firmer and of a better flavor than the linseed cake 

 butter. 



" The feeding of coconut cake to dairy cows produced a butter which gave a 

 refractometer figure below the normal, while the Polenske number was very 

 high. The Reichert-Meissl number was, however, quite normal. Feeding with 

 linseed cake influenced the butter in respect to the Polenske number, butters 

 from this source giving a figure much below the normal. The refractometer 

 figure was also influenced, a distinct rise being noticeable. Coconut cake pro- 

 duced a butter of good quality, possessing better keeping properties than butter 

 from linseed or cotton cakes. The Shrewsbury and Knapp method for detection 

 of coconut oil in butter possesses practically no advantages over the Polenske 

 method." 



The influence of feeding cotton-seed meal upon the composition of the 

 fat in butter, A. J. Swaving (Rev. G6n. Lait, 8 (1911), Nos. 20, pp. 457-^65; 

 21, pp. 486-493). — After tests of the value of Halphen's reaction with authentic 

 samples of cotton-seed oil and with mixtures of the oil and butter, the author 

 details the results of three feeding tests with cows which were conducted during 

 Ihe years 1901 and 1902. The presence of cotton-seed oil could be noted in the 

 butter fat after the first day, and it remained in the fat yielded for several 

 days after ceasing to feed the cotton-seed meal. In the fat up to 5 per cent of 

 cotton-seed oil could be detected with Halphen's reagent. By feeding cotton-seed 

 meal no influence was exerted upon the yield of fat, the Reichert-Meissl number, 

 or the refractive index. 



