,uif Substances, and on their Combinations wuh c.^^^^..^...,. .. 



it,coTitaiaed 0-0.'J2 gr. of potash : coiisequeixtiy, the marg rine ^m 

 tlie alkali of the matter precipitated must have been there in t' 

 proportion of 100 to 7*62. The analysis which i made with t' 

 muriatic acid gave me the proportion of 100 to 7'iJ>, which d 

 fers from the former only by 0-33. - If we take the medium, • 

 shall have 7*78, and we shall then find that one-eighth of t 

 alkali contained in the pearly matter was separated by thevvati 



20. The matter which was preci]jitated from the alcohol 

 the water did not yield a perceptible quantity of alkali to t 

 latter; but if we add a little hematine, a part is then separate 

 This is proved by the two following experiments. 1 boiled t 

 matter in water, divided the liquid into two equal portions, i 

 tered one of them, and afterwards put hematine into the filter 

 liquor, and also into that which was not filtered. Tiie fornv 

 makes ;;lmost no change on the colouring principle, wlicre. 

 the latter renders it purple by giving alkali to it. This decon. 

 position takes place only in consequence of the uniteil action ( 

 tlie water and heinntine ; for, if we mix the diisolved matters i 

 alcohol, the hematine undergoes no change. (18.) 



21. The matter which was precipitated from the alcohol b; 

 water, redissolved twice in the alcohol, was deposited ui)on cool- 

 ing in the state of true mother of pearl matter, which contains 

 100 of margarine and S'SS of pota^-h. 



22. It results from what I have said on the precipitation I 

 means of water from the alcoholic solution of mother of pea 

 substance: 1st, that in this precipitation there is a quantity ( 

 potash separated from the margarine, which seems to be one 

 eighth of that which is combined with it ; that this separatior 

 is owing to the afinitv of water for alcohol and the potash, and 

 to the insolubilitv of t!ie margarine in water. If boiling water 

 takes less alkali from ciie mother of pearl snb"-taiice than cold 

 water poured into the alcoholic solution of the latter, this seems 

 to ari-e, because in the second case this licpiid exercises its ac- 

 tion on a dissolved bodv : wh jrcas in the otb.er it exercises it on 

 a solid, the cohesion of which is an obstacle to its solvent power. 

 2d, That tiic matter precipitated from alcohol yields little if any 

 alkali to boiling water, but yields it to an acjueou^ solution of 

 hematine. 3dlv, That bv redissolving it in boiling alc(jhol, it is 

 precipitated in motlier of pearl matter upon cooling. This 

 proves, that in this substance the two elements exist in a pro- 

 portion in which they enjoy a cohesiim sulhcient to determine 

 the separation of the excess of margarine from which the vvatc 

 had taken the alkali. 



23. This last con>ideration led me to incjuire if marga- 

 when presented to a hot solution of potash«Jiolding a (p; 

 of alkali much more considerable than that which was u 



N4 



