,gf) WATER tKF THE TJEiU SKA. 



»oda, 54 soda, -the t>n!y cme which I shall now relate consisted in precipi* 

 *6aci(j. tating the acid by a solution of silver from a known weight 



of muriate of soda, and inferring the proportion of acid and 

 «lkali from the quantity of luna cornea obtained. This 

 however required a previous exact knowledge of the propor- 

 tions of acid and silver in luna cornea. In order to ascer- 

 tain this point, a known quantity of acid was precipitated 

 iWuHate of sil- ^Y "itrate of silver, and the weight of the luna cornea, after 

 ^c, being uieltcid and heated to redness, indicated 19*05 parts of 



1906 acid.* ^"^'^ ^^ 80*95 of oxide of silver. The composition of com- 

 mon salt, calculated from these data, proved to be 46 parts 

 of acid to 54 of soda. 



^ECT. Hi. 



Comparative Analysis of artificial Sotutions. 



Artificial soIu> ^ shalt not enter into all the particulars of the various 



tions of these analyses of artificial solutions, resembling: the water of the 



Mite analysed. ^-^ , ^, i • , t , ■,,-,., , , 



Dead Sea, which du-ected me in the choice Oi the method 



which I ultimately adopted. But it may be proper to state, 

 in a summary manner, the principal means which were tried> 

 and their respective defers and advantages. 



These Artificial mixtures all contained the three muriates 

 above mentioned, but in each of them the small quantity of 

 feelenite was altogether disregarded. 



1. The first of these solutions was evaporated to dryness, 

 and the residue exposed for near an hour to a red heat iri 

 a platina crucible pretty closely covered. The object of this 

 was to drive off the acid from the magnesia (muriate of 

 magnesia being decomposable by heat), and after separating 

 this earth from the otlier salts by means of distilled water, 

 to precipitate the lime by carbonate of amitionia, and to ob- 

 tain the muriate of soda by evaporation to dryness. But I 

 Hdat didjiot soon found, that the complete decomposition of muriate of 

 fcompol^the'^' magnesia by heat, under these circumstances, was extremely 



muriate of difficult, if not impossible; and acco-rdihgly the results ob- 

 inagnciia, tained from this method indicated considerably less magne- 

 sia and proportionally more lime, than the solution really 

 contained. The quantity of common salt was tolerably ac- 

 turate. 



2, From 



