9. ~The determination of cacao husk~, which will be for the most part a matter of ascertaining the amount of raw or crude fibre, could formerly only be effected by means of the microscope. In 1899 Filsinger[188] proposed a method of levigation which according to P. Welman’s[189] gives trustworthy results. Manifold treatises have been devoted to the subject, and it would be advisable to turn a few of these up and compare the details of the accounts.[190] In this method, which works best with the modifications suggested by Drawe (see below) 5 grammes of cocoa or chocolate are defatted with ether and dried, then ground in a mortar after a little water has been added, and levigated with about 100 ccm of water in a cylinder. The liquid is allowed to rest for some time and the suspended matter poured off almost to the sediment, which is again shaken up with fresh water, allowed to settle, and the operation repeated until all the fine particles have been floated off and the water over the sediment no longer becomes cloudy, but remains clear after the coarse and heavy particles have settled down. The powdery sediment is collected on a watch glass, dried in the water bath, and after being cooled down in a desiccator, weighed. The weighed residue is then softened with caustic soda and glycerine and examined under the microscope. The presence of any cotyledon particles must be carefully observed, such as have escaped separation in the grinding and levigation, and whether particles of husk or epidermis or germ preponderate. With proper levigation only traces of cacao substance, especially here recognisable by the cacao starch, should be present. The sand, which always adheres to the shells in the fermenting and drying operations, is also easily recognised and many indications as to the nature of the article under investigation can be noted by the use of a simple magnifying glass applied to the washed residue on the watch glass before drying. Examined in that way, a sample of so-called Cocoas from unshelled beans gave from 6 to 8 percent of husk; usually good cacao powder shows a maximum of 2·5% husk. It is true that from this Filsinger-Drawe procedure the correct percentage of shell can only be estimated in very rare instances, for when it is necessary to be absolutely fair to all concerned in the manufacture, the cacao must be so often washed until no grains of cacao starch are visible under the microscope; and so the result is often too small, more especially in the case of the finer qualities. But when all particles of starch have been removed, the finer particles of shell have often been taken along with them. Yet when the residue certainly exceeds the standard percentage of shell, it may be taken for granted that adulteration with husks has been carried to excess, or that the cleansing processes have not been effectively carried out. There is no other method which yields the same degree of certainty.