[34] Rettger, Amer. Jour. of Physiol., viii. p. 284; Koch’s Jahresbericht, 1903, p. 112. With these bodies no experiments on skin have been made, so far as I am aware, except with indol and skatol. Kathreiner found that these had a slight reducing action on skin, so that one may say they play some part in the puering. The action of the bile salts, glycocholate and taurocholate of soda, also needs investigation. These have an indirect effect in the puer, as they favour the development of some species of bacteria (chiefly _coli_) and hinder the growth of others. Some action has also been ascribed to sulphuretted hydrogen, but in the puer liquors which I have examined no H_{2}S was found, either before or after the skins were entered. It will be seen that the ammonia compounds in the bate are not of themselves particularly fitted for the purpose of removing lime[35] from the skins, but owing to bacterial action (which we shall treat of in Chapter IV.), acids are produced which combine with the ammonia, and in this way the small quantity of these compounds originally present is continually being regenerated while the bating is in progress. Ammonia is set free by the lime in the skins. It is then neutralized by acids produced by bacteria, and thus acts as a carrier for the acids, and the bate remains in a nearly neutral condition. As the lime increases in the liquid the action of the bacteria diminishes, and finally the alkalinity becomes too great to allow the bacterial or chemical action to proceed further. [35] Deliming with acids when carefully done takes out more lime than the puer, without in any way injuring the skin, but when tanned with sumac the bellies and axilla are harsh, and of a browner colour than in a puered skin. Analysis showed no more lime in the brown portions than in the centre of the skin, where the colour was quite bright and satisfactory. It will be noted that the _concentration_ of the active salts in the bate is extremely small. If the amine compounds be assumed to consist of ethylamine butyrate or lactate, the concentration of the solution is approximately 1 grm. per litre; it is important that the concentration of salts should not greatly exceed this amount. I have found by experiment with ammonium chloride solutions, that the best reducing action is provided by a concentration of 0·7 to 1 grm. NH_{4}Cl per litre; if the concentration be raised to 2 or 3 grm. per litre, the skins become “leathery” and do not fall properly. The alkalinity must not be greater than 3–5 c.c. N/10 per 100 c.c. bate, for the bate to work at its best. *Solution of Skin Substance during the Puering.*--The determination of the total skin substance dissolved by the puer is best done by Kjeldahl’s method before and after the goods.[36] [36] Vide Procter, L.I.L.B., 1908, p. 64.