Bo7inington Water, near Leith, Scotland. 309 



manganese, scarcely to be weighed. The caustic fluid* ob- 

 tained from Gwas supersaturated by hydrochloric acid, when 

 ammonia gave a precipitate (J') of O'lS* gr. of alumina. The 

 filtered fluid from the alumina was evaporated to dryness; 

 the residue, after being heated till the ammoniacal salt was 

 dissipated, left, when dissolved in water, a small trace (g) of 

 alumina, weighing 0*063 gr.f The fluid obtained from the pre- 

 cipitate (F) was neutralized by hydrochloric acid, and quickly 

 evaporated with carbonate of potassa, and when nearly dry 

 mixed with boiling water and filtered ; the precipitate (F) 

 having been properly edulcorated with boiling water, weighed 

 after ignition 10*857 grs., it was redissolved in hydrochloric 

 acid and treated with hydrosulphate of ammonia ; the precipi- 

 tate hereby obtained was collected upon a filter, and gave after 

 a similar treatment as r, O'OiS gr. oxide of manganese (h); 

 deducting this amount from 10-8.57, there remains 10*809 grs. 

 of magnesia (0 J, equal to 22*372 grs. of carbonate of mag- 

 nesia. The fluid obtained from the precipitate (F) still con- 

 tained some traces of magnesia, it was therefore precipitated 

 by ammonia and phosphate of ammonia, when the precipitate, 

 after lixiviation with dilute ammonia and ignition, gave 1*127 

 gr. of phosphate of magnesia, equivalent to 0*413 magnesia 

 (/t), equivalent to 0*854 gr. carbonate of magnesia. The 

 ammoniacal fluid obtained from the precipitate (E) was next 

 precipitated by oxalate of ammonia, and as soon as the pre- 

 cipitate had settled it was quickly filtered and lixiviated §. 

 This precipitate (I) was ignited under the well-known pre- 



* As caustic potassa or soda is seldom if ever quite free from traces of 

 silica or alumina, I ascertain, previous to the application of these tests, the 

 quantity of the latter, for which a due deduction is made. 



+ I have here to observe, that every ingredient obtained was, after its 

 weight had been ascertained, again examined for its purity. The alumina, 

 as well as the oxide of iron, was examined before the blowpipe for phos- 

 phoric acid ; besides, these precipitates were dissolved in hydrochloric acid, 

 precipitated by ammonia, and re-dissolved by a gentle heat in acetic acid, 

 when after the lapse of one day no sediment had formed, evidencing the 

 •absence of phosphoric acid. The caustic fluid from which the last trace of 

 alumina was separated was likewise tested for phosphoric acid, by adding to 

 the fluid, neutralized by hydrochloric acid, chloride of calcium j but no trace 

 of it was discernible. 



X It is very difficult to separate the last traces of magnesia by carbonate 

 of potassa; the fluid should be boiled quickly to dryness, and the residue 

 lixiviated by boiling water. The fluid obtained from the residue should 

 invariably be examined if it is free from magnesia. As in the above process 

 the fluid containing the magnesia was not boiled to dryness, it accounts for 

 the larger amount of magnesia obtained afterwards. 



§ When magnesia and lime are to be separated, both in considerable 

 quantity, it is requisite to have the menstruum well-diluted, and to filter 

 the oxalate of lime as soon as possible without interruption, so as to pre- 

 vent the admixture of oxalate of magnesia. 



