290 MESSRS. THORPE AND MORTON ON THE 



a temperature of about i8o° C._, until its weight appeared 

 to be constant. Sea-water contains a notable quantity of 

 magnesium chloride, which decomposes during the pro- 

 cess of drying, evolving hydrochloric acid. This source of 

 error was easily obviated by adding, as recommended by 

 Mohr, a small known quantity of recently ignited sodium 

 carbonate to the water, when the magnesium chloride is 

 converted into the more stable carbonate. 









Amount in 1000 





Water taken. 



Residue obtained. 



grams of water. 





grams. 



grams. 



grams. 



I. . 



■• 5i"i564 



17309 



33*8360 



I. . 



507608 



17178 



33-8411 



Mean 33*83855 



According to Forchhammer, the amount of the total 

 saline constituents of the water of the northern portion of 

 the Atlantic, between the parallels 30° N. lat. and a line 

 from the north of Scotland to the north of Newfoundland, 

 is subject to very slight variation : the mean quantity 

 amounts to 35*976 grams per 1000. It is therefore evi- 

 dent from the above result, that the relative amount of 

 solid matter contained in sea- water is perceptibly dimin- 

 ished in the neighbourhood of coasts ; and this in the case 

 of the Irish Channel can only arise from the admixture of 

 fresh water flowing down in the form of rivers from the 

 land. This conclusion is fully borne out by the older 

 analyses of Clemm, Figuier and Mialhe, and Bischof, of 

 the water of the German Ocean, collected in the neighbour- 

 hood of the coasts. In no case did the amount of saline 

 constituents exceed 33 grams per 1000. (Min. 30*5, max. 

 32-8.) 



