Upon the whole we concluded that this method, which has been 

 considered by several chemists of note as the best rapid test yet offered, 

 although quite reliable in the hands of a careful worker in a chemical 

 laboratory, is not suitable in the hands of the average man. Not to 

 our knowledge has it ever been used or recommended for work with 

 curd. We tested it with curd and found it to work fairly well, but of 

 course the above-mentioned objections, with the exception of number 

 4, still hold here. Several modifications of the beaker method have 

 appeared from time to time. 



Gray's test, consisting of heating a ten gram sample of butter with 

 a mixture of am)l acetate and amyl valerianate in a glass flask, to which 

 are attached a ofass condenser and graduated receiver reading in per- 

 centages, presented none of the above difficulties ; but we found 

 several objections to it of a much more serious nature. It gave, with 

 us, readings altogether false as compared with the official method, con- 

 cerning which there is no doubt whatever as to its reliability. Further, 

 so much care is required to prevent loss of moisture through the open 

 graduated neck and to guard against the breaking of a comparatively 

 expensive and fragile piece of glassware, that these points themselves 

 would tend to militate against its use in a creamery. In addition, the 

 apparatus requires putting the different portions together and taking 

 them apart again for each test. Lastly, the reagent used is rather expen- 

 sive for practical use, since it cannot be used over again. 



Since Gray's form of apparatus appeared on the market a so-called 

 "improved form" of it has been brought forward. We tested this, also, 

 but found it more fragile, rather cumbersome, and requiring consider- 

 able time and trouble to clean, although it gave more accurate readings 

 than the original form. This greater degree of accuracy is due to a -modi- 

 fied form of graduation of the scale. The errors in the results with the 

 original form of Gray's test are due, chiefly, to absorption of moisture 

 by the reagent used and not to any lack of condensation, providing suffi- 

 cient care is taken in conducting the test. 



Being dissatisfied with all the methods investigated by us, we under- 

 took to work out a form of moisture test that would prove more satis- 

 factory. We decided that the following requisites were necessary, and 

 we kept them continually in view during our work : 



(i) A rapid and reasonably accurate method. 



{2) An inexpensive form of apparatus, both to purchase and to use. 



(3) A durable form of apparatus and one easy to clean. 



(4) A method that requires no great amount of attention or care 

 in operating, and that can thus be used by the average factory man. 



(5) An apparatus than can, if damaged, be easily repaired at small 

 cost. 



(6) A method equally suitable for testing butter, curd and cheese. 



