It is convenient to determine the water in the sample subsequently to be used for the gravimetric determination of the fat, and this is secured by the adoption of the paper coil method, as suggested by the author, or by the use of a perforated metal tube containing porous asbestos, as proposed by Babcock.[423]
The process is conveniently carried out as follows:
Provide a hollow cylinder of perforated sheet metal sixty millimeters long and twenty millimeters in diameter, closed five millimeters from one end by a disk of the same material. The perforations should be about 0.7 millimeter in diameter and as close together as possible. Fill loosely with from one and a half to two and a half grams of dry woolly asbestos and weigh. Introduce a weighed quantity of milk (about five grams). Dry at 100° for four hours. During the first part of the drying the door of the oven should be left partly open to allow escape of moisture. Cool in a desiccator and weigh. Repeat the drying until the weight remains constant. Place in an extractor and treat with anhydrous ether for two hours. Evaporate the ether and dry the fat at 100°. The extracted fat is weighed and the number thus obtained may be checked by drying and weighing the cylinder containing the residue.
The asbestos best suited for use in this process should be of a woolly nature, quite absorbent, and, previous to use, be ignited to free it of moisture and organic matter. A variety of serpentine, crysolite is sometimes used instead of asbestos. When the content of water alone is desired, it is accurately determined by drying in vacuo over pumice stone ([page 33]).
The methods above mentioned are typical and will prove a sufficient guide for conducting the desiccation, either as described or by any modification of the methods which may be preferred.
448. Calculation of Total Solids.—By reason of the ease and celerity with which the density of a milk and its content of fat can be obtained, analysts have found it convenient to calculate the percentage of total solids instead of determining it directly. This is accomplished by arbitrary formulas based on the data of numerous analyses. These formulas give satisfactory results when the samples do not vary widely from the normal and may be used with advantage in most cases.
Among the earliest formulas for the calculation may be mentioned those of Fleischmann and Morgen,[424] Behrend and Morgen,[425] Claus, Stutzer and Meyer,[426] Hehner,[427] and Hehner and Richmond.[428] Without doing more than citing these papers it will be sufficient here to give the formulas as corrected by the most recent experience.
In the formula worked out by Babcock the specific gravity of the sample is represented by S, the fat by F, and the solids not fat by t. The formula is written as follows:[429]
| t = ( | 100S - FS | ) (250 - 2.5 F). |
| 100 - 1.0753FS |
In this formula it is assumed that the difference between the specific gravity of the milk serum and that of water is directly proportional to the per cent of solids in the serum, but this assumption is not strictly correct. Even in extreme cases, however, the error does not amount to more than 0.05 per cent.