[ 410 ] 



LXVI. On the Cartesian Barometer: 

 By William Roxburgh, M.D* 



SOON after the discovery of the variations in height of the 

 barometer, Descartes proposed the following mode of ren- 

 dering them more conspicuous, almost as much so as they are 

 in one filled with water alone. He suggested that two tubes 

 should be joined to the opposite ends of a short wide cylinder so 

 as to form one straight tube, which, being closed at one end, 

 was to be filled with pure water and mercury in such proportions 

 as to allow of the two fluids at all pressui-es meeting in the 

 cylinder. In this, the Cartesian barometer, the pressure of the 

 atmosphere is balanced by the water and mercury conjointly, but 

 the variations of pressure' are indicated chiefly by movements of 

 the water, as the level of the mercury varies little in consequence 

 of the large area of the cylinder. The movements of the water 

 and mercury are to each other inversely as the areas of the tube 

 and cylinder. The scale is that of the common barometer en- 

 larged as in the wheel barometer; when, therefore, the move- 

 ments are said to amount to so many hundredths of an inch, it 

 is to be understood as meaning that they are equal in value to 

 that height of mercury. The scale can be enlarged so as to 

 render movements of 4^0^*^ ^^ ^^ ™^^ visible to the unassisted 

 eye. 



The only records of this instrument that I have seen, state 

 that the air contained in the water is given ofi" when the pressure 

 is removed, and so renders its indications incorrect ; also that 

 this imperfection is irremediable. This depression, amounting 

 in one year, in my first experiment, to only -02 of an inch, has 

 led me to suppose that the depression which caused the plan to 

 be set aside was owing to the force of vapour, which was not so 

 well understood at that time as at present ; and as many varia- 

 tions of pressure are easily seen in this barometer, which would 

 escape notice in the mercui-ial one, and if not attended to give 

 rise to error, I think it \n\\ prove a valuable addition to a stand- 

 ard barometer, though never a substitute for one. 



In hopes of getting rid of the air, and of lessening the cor- 

 rection required for the force of vapour, I tried several fluids 

 in place of pure water. Among these was oil of turpentine ; 

 this caused a rapid evolution of gas and blackening of the mer- 

 cury, and depressing the column several inches in a few minutes. 

 A saturated solution of muriate of soda seemed at first more suc- 

 cessful, but in a short time the column became depressed, and 

 this depression continuing to increase at a regular rate, the tube 



* Communicated by the Author. 



