The difference in the total nitrogen found multiplied by 5·6 gives the amount of skin substance dissolved by the bate, assuming the amount of nitrogen in the dry ash-free skin to be 17·8 per cent. If very great accuracy be required, a small correction for nitrogen, brought into solution from the puer itself, is necessary.[37] This correction must be ascertained for the particular puer used by actual experiment. [37] In a blank experiment to determine this, 0·0476 grm. N per litre was found to be brought into solution from the puer. The following figures give the results in grams per litre obtained in the puering of sheep grains. 50 c.c. of the filtered puer liquor are slightly acidified with sulphuric acid, evaporated nearly to dryness and Kjeldahled in the usual way. Mean Total nitrogen before skins {0·2632} 0·2604 {0·2576} " " after " {0·4928} 0·4844 {0·4760} Difference -- 0·2240 Equivalent to 1·254 grm. skin substance per litre. This was somewhat over a kilogram of skin substance for the paddle in question, and equal to 1 per cent. of the dry ash-free skin. As to the differentiation of the dissolved skin substance into albumoses, peptones, monamino acids, diamino acids, ammonia, etc., a modification of Stiasny’s method[38] for the examination of soaks and old limes may be used. [38] Stiasny, On Old Limes. Collegium 1910, p. 181. The method is based on the fact, discovered by Schiff, that formaldehyde reacts with amino acids, forming methylen-amino acids, which are distinctly acid and allow a sharp titration with phenolphthalein as indicator, while the amino acids themselves react almost neutral. Soerensen has worked out a method on this basis for the determination of different amino acids, and for tracing the course of hydrolysis of albuminous matters. Instead of using phenolphthalein as an indicator, the electrometric apparatus of Sand (see p. 76) is employed. 50 c.c. of the filtered puer liquor are put into a beaker, the hydrogen electrode is immersed in the liquor, and the potential difference (P.D.) observed; this gives the hydrogen ion concentration of the solution. 10 c.c. of neutral formaldehyde solution (40 per cent.) are added, and the P.D. again observed; it will be found to diminish rapidly, but soon becomes constant, indicating that the reaction is a quick one. The increase of acidity, as shown by the lowering of the potential difference, is due to the acidity developed by the combination of the formaldehyde with the amino acids forming methylen-amino acids of appreciable hydrion concentration. The amount of such acids is estimated by titrating with N/10 caustic soda solution until the P.D. rises to the same voltage as that originally found. The following figures were found in an experiment:--