278 American Association. 



nomenon. After some consideration I adopted the following 

 method, as one sufficiently trustworlhy to afford an answer to tbe 

 question under review: — I procured a leaden pipe one inch in 

 diameter, and bent into the form of the letter U : each upright being 

 about thirty-nine inches in height, and tlie connecting piece at 

 the bottom rather more than twelve inches long. This I filled up 

 to about an inch on each side with a solution of common salt 

 in rain water (the salt being present to the amount of 3.786 per 

 cent.,) and then I carefully closed one end, leaving the other end 

 open, but protected from dust by a cone of silver-paper fixed on a 

 bent wire, and so arranged as not to prevent evaporation. Tl)e 

 pel' centage of salt (3, 786) was carefully ascertained, and the appa- 

 ratus left in an unoccupied room, the window and door of which 

 were kept almost constantly open, in order to promote the evapora- 

 tion of the solution as much as possible. After the lapse of about 

 three months, (April 18 to July 14,) portions were taken from 

 each end of the tube, and from the connecting piece below, (a 

 small orifice being made in this ;) and the amount of salt in each 

 portion was accurately determined. Now if the principle of dif- 

 fiision had not been brought into play, it is evident that the solu- 

 tion in the open limb of the tube ought to have been stronger 

 than that in the closed limb, although, by the circulating process, 

 the amount of salt at the top and bottom of the former might have 

 been alike ; and, again, it will be equally evident that if the prin- 

 ciple of diffusion were brought into play, the supposed sinking of 

 the surface solution, as the result of evaporation, must be altogether 

 imaginary. Six separate determinations, two from each of the 

 three portions of the tube, shewed a per centage of salt essentially 

 the same. The following table exhibits the results obtained : 



( A. From the top of the open limb,. 



1 } B. From the bottom of the same,. 

 I C. From the top of the closed limb, 

 C A. From the top of the open limb,. . 



2 ^ B. From the bottom of the same,. . 

 ( C. From the top of the closed limb, . 



These experiments justify us, I think, in assuming, that owing 

 to difi"usion, the surface waters of the sea do not become heavier 

 than the lower strata simply by losing water by evaporation. It 

 is quite true, that under the influence of evaporation a lowering 

 of temperature may take place, and that an upward and down- 

 ward circulation, to a certain extent, may in this manner be pro- 



Solution. 



Am. of Salt. 



Per ct'ge 

 of salt. 



302-261.. 



. 11-59... 



3-030.. 



300-24.., 



. 11-51... 



3-835.. 



288-60... 



, 11-055.. 



2-831.. 



264-84.., 



. 10-16... 



3-837.. 



290-10... 



, 11-12... 



3-833.. 



306-66... 



ll-YS... 



3-832.. 



