808 EXPERIMENT STATION EECOED. 



monia in milk. From their work tliey conclude that the presence of ammonia 

 in milk is always an indication that it has been kept under unhygienic condi- 

 tions or has been watered. On the other hand, they show that the absence of 

 ammonia in milk does not always indicate that the milk has not been watered 

 or has been properly cared for. 



Commercial tests for m.oisture in butter, C. W. Larson (Pennsylvania 

 8ta. Rpt. 1910, pp. lJi7-155, pis. 5).— This is a study of the relative value of the 

 several commercial moisture tests for butter thus far proposed (the Wagner- 

 Gray, Irish, Wisconsin-high-pressure-oven, Ames-paraffin, and Marshall tests) 

 as compared with the official gravimetric method. 



" The results obtained from Wagner's moisture test are not always accurate. 

 Very accurate results can be obtained with the Wisconsin high pressure oven, 

 the Ames, and the Marshall tests, and if care is taken in the heating of the 

 cup in the Irish test good results can be secured." . . . 



Observations on some m^ethods of estimating coconut oil and butter in 

 butter and margarin, C. Revis and E. R. Bolton {Amilyst, 36 (1911), No. 

 424, pp. 333-3Ji2, fig. 1). — The Polenske method according to these authors, who 

 have applied it to numerous samples, is the only chemical method whereby 

 small amounts of coconut oil can be detected and approximately estimated in 

 butter, failures being due largely to a lack of scrupulous attention to the 

 experimental details involved in the method. 



The authors have, however, compared the effect of maintaining a constant 

 temperature in the still-head with results obtained under the usual conditions 

 of work. They conclude that " so long as reasonable care is taken to protect 

 the still-head from draft, no error arises from this source; and as the mainte- 

 nance of a high temperature depresses both the Reichert-Meissl and Polenske 

 values, there is no advantage in employing a specially protected high-tempera- 

 ture still-head, beyond the fact that time, and to a certain extent the type of 

 distillation, become of less importance." 



The Shrewsbury and Knapp method (E. S. R., 24, p. 515), the cadmium 

 method of Paal and Amberger (E. S. R., 21, p. 110), and Kirschner's method 

 (E. S. R., 16, p. 1050) were also studied. The Shrewsbury and Knapp method 

 was open to the objection that concordant results could not be obtained with the 

 same samples. It was also modified by the authors, but the figures obtained with 

 it were not as reliable as those obtained by the Polenske method. The cadmium 

 method after a few tests was abandoned on account of its laborious execution 

 and bad results in the authors' hands. 



The Kirschuer method was found to offer little in advance of the Polenske 

 method for coconut oil and butter, but it had a decided advantage when applied 

 to margarin containing both butter and coconut oil, or margarin containing 

 coconut oil alone. It is the authors' conviction " that it is the only published 

 method at present available which will decide whether butter fat is present 

 in a margarin containing coconut oil, even if the amount be only 2 per cent in 

 conjunction with 75 per cent or more of coconut oil." The method as employed 

 by the authors is described in detail. 



Examination of Shrewsbury and Knapp's process for the estimation of 

 coconut oil, R. Ross. J. Race, and F. Maitdsley (Anahjfft, 36 (1911), No. Ji22, 

 pp. 195-198). — This investigation was conducted with Irish, Danish, Finnish, 

 Siberian, German, and English butters, coconut oils from various sources, cacao 

 butter, and coconut stearin. 



The highest Shrewsbury and Knapp figure for the butter was found to be 

 35.5 in contrast to 32 as suggested by Shrewsbury and Knapp as the maximum. 



The authors point out that " in considering the genuineness or otherwise of 

 a butter, it is doubtless advisable not to rely entirely on any particular figure. 



