8 EXPERIMENT STATION RECORD. 



that drying in a vacuum over sulphuric acid is the most satisfactory procedure in 

 estimating water in foods and physiological products. 



The high vacuum which is needed, the authors found, could be obtained by intro- 

 ducing a small quantity of ether into the desiccator containing the samples and 

 exhausting with a suction pump. It is desirable, in the authors' opinion, to have a 

 manometer in the desiccator and to dry the samples in aluminum dishes of special 

 construction with covers, although suitable glass dishes with covers may also be used. 



As regards the time required for drying, "a marked difference exists between 

 materials of a vegetable and those of an animal nature, the former retaining their 

 moisture persistently, while the latter have, in general, lost all but mere traces of 

 their moisture after 2 weeks in a high vacuum. For animal materials, therefore, we 

 may say that 2 weeks in a high vacuum suffices to remove practically all moisture. 

 For the most accurate work a longer time is to be recommended, though the slight 

 loss in weight subsequent to 2 weeks' desiccation can hardly effect ordinary physio- 

 logical or chemical research. With vegetable materials, desiccation should proceed 

 for a period longer than 2 weeks [as slight loss in weight was noted after this]. 



"As a matter of fact, however, it can readily be seen that the sample, as soon as 

 prepared for analysis, can be placed in the vacuum and weighed as soon after 2 

 weeks as the results are needed. In general, the time between the preparation of a 

 sample for analysis and the time when the results are imperatively needed is con- 

 siderably over 2 weeks. For preliminary determinations in metabolism experiments, 

 where an approximate knowledge of the water-content of foods is necessary, the 

 usual five-hour heating will suffice, and the final calculations may be based on the 

 determination by the vacuum method. For technical work, obviously the method 

 is, as already stated, too time-consuming, though the advantages may appeal to the 

 technical analyst and the method may possibly find some use in special cases. . . . 



"The complete dehydration of butter in a high vacuum is of especial interest, 

 when we consider the difficulties usually experienced in securing an equal distribu- 

 tion of the material in the sample dish and the passage of the water through the 

 supernatant layer of melted fat. It is worthy of note that in several tests made with 

 the vacuum method no special degree of importance could be ascribed to an even 

 distribution of the sample. Desiccation was equally complete whether the sample 

 was first melted in a thin layer on the bottom of the dish, or large irregular lumps 

 of butter were used." 



Studies were made of the effect of different degrees of rarefaction on the dehydra- 

 tion of samples. When the atmospheric pressure was reduced to <S6 mm. the effi- 

 ciency of dehydration was considerably increased over that noted in the desiccator 

 at atmospheric pressure, and when the degree of rarefaction was less than 1 mm. of 

 mercury the increased efficiency of desiccation became very marked. "This is easily 

 explained by the fact that the tension of aqueous vapor at the laboratory tempera- 

 ture (20° C. ) is about 17 mm. of mercury, and consequently a pressure of 86 mm. 

 requires a considerably higher temperature to cause the tension of aqueous vapor to 

 equal that of the gaseous medium. On the other hand, with the high vacuum, the 

 temperature of the room is perhaps 20° above the boiling point of water at this degree 

 of rarefaction, and consequently evaporation of water proceeds with great rapidity. 

 A consideration of these facts explains the slow action of vacuum-drying as usually 

 carried out, and also accentuates the importance of using a manometer inside the 

 desiccator to indicate the pressure conditions after exhaustion." 



In view of the results obtained the authors do not believe that drying in partial 

 vacuum with heat is necessary or desirable. 



The determination of water in molasses, G. Testoni (Staz. Sper. Agr. ItaL, 

 37 {1904), No. 4-5, pp. §QG-$69), — Different methods of estimating water in molasses 

 were compared. 



