FREEZING POINT. 81 



Recknagel's phenomenon appears to be unconnected with the 

 milk-sugar, and Recknagel's explanation is not the correct one. 

 It is difficult to reconcile the idea that it is enzymic, with the 

 fact that the rise is more rapid at low than high temperatures. 



The author's experiments on the change of density and specific 

 heat of cream by heating made in conjunction with S. 0. Rich- 

 mond have shown that Recknagel's phenomenon is due largely 

 to the increase of density of the fat on solidification. 



Contrary to the author's former conjecture, there seem to be 

 no particular periods of the year in which Recknagel's pheno- 

 menon is observed or not. Samples have been found at all 

 seasons which show a marked change in specific gravity, 

 while others examined almost simultaneously have shown no 

 change. 



It must be mentioned that Recknagel's phenomenon has been 

 denied by some. Smetham attributes the change in specific 

 gravity solely to the presence of air bubbles. The weight of 

 evidence is, however, greatly against this view ; it is incon- 

 ceivable that air bubbles generated by milking a cow should be 

 persistent for twelve hours, while if they are formed in the milk 

 by other means, say by running through a separator, they dis- 

 appear in one hour. 



The final specific gravity is always taken as the true specific 

 gravity of milk, and the term is so used in this volume. 



Freezing Point. During the past twenty years a large number 

 of investigations on the freezing point of milk have been made ; 

 and most observers have agreed that it is a very constant figure, 

 but there have been distinct differences between the published 

 results of different observers. The freezing point has been stated 

 at from 0-51 to 0-59, though each observer has usually 

 found the variations not to exceed 0-03 in the milks that he 

 has examined, and tht differences have been due to the experi- 

 mental errors, usually fairly constant for each observer, of the 

 method. The method has acquired some importance, as not 

 only is there distinct evidence that milks low in solids not fat 

 have normal freezing points, but the method has been made 

 official in Holland and Queensland. Monier-Williams, in a 

 comprehensive report to the Local Government Board, has 

 reviewed the subject, and pointed out the many precautions 

 that must be taken to obtain accurate results, and shown 

 that numerous corrections must be applied to the observed 

 results, and that most of the recorded figures are appreciably 

 too large. 



The errors are : 



(1) Those due to the thermometer, which may be eliminated 

 by taking the freezing point of dilute solutions of cane sugar of 



6 



