It is necessary that the balance used should weigh accurately to milligrams; and it must carry at least 50 grm. on each pan; while it is more convenient that it should carry 100 or more, it is always possible with a little ingenuity, to manage within 50 grm.; and if a cheap balance must be used, the smaller size will probably be more accurate. Balances of this sort can now be got for two or three pounds, though it is in all respects better to obtain one of first rate quality, which should cost about ten pounds. The balances of Verbeek and Peckholdt, of Dresden, from their simplicity and rapidity of weighing, have given great satisfaction in technical work in the Yorkshire College. Whatever economy be exercised in the choice of the balance, it is essential that the set of weights should be of the greatest accuracy, and especially that all the weights of one denomination (10 grm., 1 grm., etc.), should accurately balance each other. Even after all precautions are taken, it is desirable that those weights which are in duplicate should receive distinguishing marks (e.g. with a centre-punch), and should always be placed on the scale in the same order; and, not only on account of possible inaccuracy, but to save time, it is desirable to reject basins which are so nearly of an even weight (20, 25, 30 grm.) that when weighed with the residue (0·3-0·4 grm.), a change of the larger weights may be required, since it is to be remembered that any error of the weights employed is concentrated on the small weight of the residue. After deducting the weight of the empty basins, the weight _in milligrams_ of the two residues of 50 c.c., which should be practically alike, are added together, and the sum divided by the weight _in grams_ of tanning material used; which gives the percentage of “total soluble matter.” [Illustration: FIG. 66. Hide-powder Filter.] _Non-tannins._--It is now necessary to determine the proportion of the “total soluble” which consists of “non-tanning matters,” that is, of substances not removed from the solution by treatment with hide-powder. The so-called “tanning matters” removed, include colouring matters and some other substances, which though absorbed by hide, are certainly not tannins in a strictly chemical sense. (See note, p. 480.)