On the Sallness of the Atlantic Ocean. 291 



frame, and I take care to prevent the larger pictures from swag- 

 ging, by not allowing the canvass to be tightened afterwards, but 

 cutting the jiicture off the frame on which it was painted, and 

 then pasting it to very tight canvass protected from swagging by 

 pannels. This mode of preventing cracking is of real advantage; 

 for all pictures will imbibe the smoke of London, and become 

 tinged in proportion to the softness which is had recourse to for 

 this purpose. 



A third varnish is ciuickly made by oil of spike-lavender, which 

 is very good for drawings or prints, but will not do for pictures, 

 as it will dissolve the paint underneath and run down while dry- 

 ing. I remain, sir. Your most obedient servant, 

 42 Newman-street, May 17, 1816. CORNELIUS VarleY. 



L. On the Saltness of the Atlantic Ocean. By M. Gay- 



LUSSAC*. 



i\x. Lamarche, a distinguished officer of marine, collected the 

 water of the sea at different latitudes during a voyage from Rio 

 Janeiro to France in the year 1S16, and on his return to Paris 

 he brought it to me in order to submit it to some experiments. 

 The water had been taken from the surface of the sea, and pre- 

 served in glass bottles closed with stop corks, and for the most 

 part tarred. 



I at first proposed to myself to examine the nature and the 

 proportions of the sahne substances which they contained ; but I 

 found myself relieved from the necessity of such examination, by 

 the analysis of the water of the Frith of Forth by Mr. John Mur- 

 ray, who appears to have made it with much care, and from which 

 it results that this water contains in 100 parts 



Muriate of Soda 2-lSO 



Magnesia 0-486 



Lime 0-078 



Sulphate of Soda 0-350 



I contented mvself with determining the specific weight and 

 quantity of the saline matters of each specimen. The experiments 

 were made in my laboratory by M. Despretz with all possible 

 care. The specific weight was taken, by weighing successively 

 the same vessel, first empty, then full of distilled water, and after- 

 wards full of sea- water at the constant temperature of 8^ 



The total quantity of saline matters may be determined by 

 analysis, in the manner adopted by Mr. Murray; but it is more 

 simple and more exact to determine it by evaporation. This ope- 

 ration is done very conveniently by putting the water in a phial, 



* From the AnnuleH de Cli'mie for December 1817- 



T 2 and 



