in the ProduBion of Alum, g^ 



(cauftic) ammoniac, on the contrary, threw down the argillaceous earth they contained. 

 And when the lye was ftill farther evaporated, it coalefced upon refrigeration into a 

 coherent faline mafs, in which, however,the needle-fliaped texture could be diftinguifhed. 

 In another experiment, in wMch I added a little more of (the fame) foda in fuch proportion 

 that the quantity of precipitated argil became fomewhat more confiderable, I obtained by 

 evaporating the filtered lixivium to a fmall remainder, and as it cooled, longifh, flat, tabular 

 Cx-fided cryftals ; (that is, very low hexagonal prifms, with broad ends and very low lateral 

 facets, the two oppofite ones of which were longer, and the other four (horter.) After the 

 mother-water, which in proportion to the clofely grouped cryftals was but fmall in quantity, 

 had been decanted, and the cryftals waflied with a little cold water, I attempted to take them 

 out of the vefiel by means of a filver fpoon ; but they were too foft, and were immediately 

 comprefled. Cold water (at 60° Farenh.) eafily and quickly diflblved them ; I have not 

 yet determined the quantity required for that purpofe ; but certainly no more than three 

 parts are neceffary for one of the fait. The folution tafted like that of alum ; it gave to 

 paper, ftained with Jitmus, of a faint red; (the undiminiflied folution, before the (hooting 

 of the cryftals, afforded a ftronger red) ; whence it follows, that there was a fmaller pro- 

 portion of the acid in the cryftallized fait than in the whole fluid. 



My attempts to produce alum by the acid of foda, being thus fruftrated, I doubted very 

 much whether ammoniac, the properties of which is far lefs analogous to thofe of pot-aflij 

 than the properties of foda, would ferve to produce alum, by combination with fulphuric 

 acid and argillaceous earth. However, experience teaches that in this as well as in 

 various other cafes, reafonings from analogy prove very often incorreft, when reforted to 

 in our enquiries into nature. Precipitate the argil from an aqueous folution of Roman 

 alum, by means of ammoniac, wafh the precipitate well ; diflblve it then in pure fulphuric 

 acid, in the manner pointed out above ; and then continue Adding ammoniac (either car- 

 bonated or cauftic) by degrees, until the folution begins to aiTume a permanent turbidnefs. 

 Let the whole then be reduced by evaporation fo much, that for o«^ part of the diflblved fait 

 (the quantity of which may be afcertained by the quantity of alum employed at the 

 beginning of the procefs) there fhall remain from fix to eight parts of water. As this lye 

 CDols, a fait will cryftallize, which, with refped to its o£iahedral form, (in the beft way to 

 be perceived around the threads hung into the liquid) to the folidity and hardnefs of its 

 cryftals, and to its difficult folubility, &c. perfedly agrees with the alum, which is formed 

 by means of pot-afli *. 



• On this fubjeft fee this Journal, I. 318. 



B>epermefit£ 



