EXPANSION. 



273 



His figures are as follow : 



TABLE XLIX. 



Mode of Expressing Results. These figures show clearly that 

 as the temperature rises the densities of butter and mar- 

 garine tend to approach one another ; the widest difference 

 occurs at 35 C. ; he, therefore, recommends that this tem- 

 perature be adopted as the temperature at which the densities 

 of butter should be determined. 



In England a large number of determinations have been made 

 by J. Bell, Allen, Muter, and others at a temperature of 100 F. 

 (37-8 C.), and this temperature is very near that found by 

 Skalweit to give the largest difference. 



In America the temperature of 40 C. is used to a considerable 

 extent, and the author has taken a large number of densities at 

 39-5 C. (owing to the use of a thermometer which read 0-5 too 

 high). 



Estcourt proposed using the temperature of boiling water 

 (which he found to raise the butter fat to 97-8 C. [208 F.]), 

 as being easily attained. Allen and others have recommended 

 this temperature, and find no difficulty in bringing the tempera- 

 ture up to 99 C. 



There is a certain amount of confusion as to the manner in 

 which densities are expressed. To ascertain the true density, 

 the weight of a certain volume of fat should be divided by the 

 weight of the same volume of water at the same temperature 

 and multiplied by the density of water at that temperature. This 

 is very rarely done, so that few published figures are true densities. 



Muter gives the term " actual density " to the weight of a 

 certain volume of fat divided by the weight of the same volume 

 of water at the same temperature ; densities expressed thus 



nrj oo 



are usually denoted by the symbol D -^-^ for density at 37-8, 



35 oT'o 



or D j for density at 35, and the true density is often ex- 



~p _37-8 _35 

 pressed as D r^- or D -^ . 



18 



