To avoid the uncertainty involved in comparing tannin (which reduces different quantities of permanganate according to the method of titration) with so dissimilar a reducing agent as oxalic acid, it is recommended to employ tannin, titrated under precisely the same conditions as the tanning material, so that whatever method be employed, the differences will be common to both, and will so be eliminated. Prof. von Schroeder has shown (Report, p. 45) by careful experiment, that with the purest samples of tannin the permanganate value estimated on the total dry substance of the tannin varied by very little from that of the part of the tannin absorbed by hide as determined by Hammer's process, but on the average bore the proportion of 1 : 1·05. The percentage of water in an air-dried tannin must be estimated by drying a portion at 201°-212° F. (94°-100° C.) and determining the loss, and a quantity equivalent to 2 _grm._ must be dissolved in 1 _litre_ of water and 10 _c.c._ titrated with indigo in the usual way. If the permanganate value thus obtained be multiplied by 1·05, it will be equivalent to that of 2 _grm._ of chemically pure tannin. It is only necessary to determine the moisture occasionally, if the tannin be kept in a well-closed box or bottle. To ascertain if a tannin is pure enough for this, use a solution made as above described (it is not necessary to determine the moisture) and 10 _c.c._ are titrated with permanganate in the usual way. 50 _c.c._ are then digested in the cold with 3 _grm._ hide-powder (previously moistened with distilled water and well squeezed in linen) for 18-20 hours, with frequent shaking, filtered, and 10 _c.c._ again titrated. If the second titration ("not-tannin") does not exceed 5 per cent. of the total, it is _good_, but it may be _used_ so long as the "not-tannin" does not exceed 10 per cent.[P] The purest tannin examined by Prof. von Schroeder was Schering's Phar. Ger., which may be obtained of Messrs. Mawson and Swan. [Footnote P: Gallic acid suggests itself to the writer as being a good standard, since it behaves with permanganate like tannin, and being crystalline is easily purified and of definite composition.] [Illustration: Fig. 20.] The course of analysis is as follows:--