VOL. XXXI.] PHILOSOPHICAL TKANSACTIONS. 4^7 



the former, not to be easily removed without great danger of disorder and 

 breaking, by reason of the smallness of the tube in which the spirit was to 

 rise and fall. 



This was succeeded by Dr. Hook's marine barometer, made of two thermo- 

 meters, the one the common sealed weather-glass, having no communication 

 with the outward air, where the temper, as to heat and cold, was shown by 

 the swelling or shrinking of the included spirit ; the other the old thermome- 

 ter made with an inverted bolt-head, having in the globular part included air 

 somewhat rarer than the ambient, so as to make the liquor, which was to rise 

 and fall in the shank of the bolt- head, always to stand above the surface of the 

 cistern, into which its end was immersed. This showed the heat of the air by 

 its own dilatation ; but at the same time the different pressure of the atmos- 

 phere mixed with it, so that the graduation of these two thermometers being 

 adjusted to any given height of the Mercury, they would at all times, when 

 the Mercury was at that height, both show the same degree of heat : but at 

 other times, when the weight of the air was different, that difference would 

 show itself by the disagreement of the degree of heat indicated by them. 

 This will be better understood from N° 269, of the Transactions, where I 

 have described this instrument at large. This, though of admirable use at 

 sea, to give timely notice of approaching bad weather, labours under the 

 objection that it supposes the tubes of the thermometers to be exact cy- 

 linders, or of equal diameters throughout; and also that, on account of heat 

 and cold, the air and spirit have a proportional dilatation and contraction; 

 the first of which I take to be very hard to be found in ordinary glass canes, 

 and the other I fear still wants to be made out by authentic experiments. 



The last contrivance for this purpose, is that of Mr. Patrick, who stiles him- 

 self the Torricellian operator, by filling a small glass cane about 5 feet long, 

 and somewhat, but as little as may be, tapering upwards toward the close end 

 of the cane; then inverting it, without a stagnant cistern of Mercury, so much 

 of the Mercury, as exceeds the length of the column the atmosphere can then 

 support, will drop off, and leave its length equal to the then present height of 

 the common barometer. Now when the barometer rises, this length in the 

 cane becomes greater by the Mercury's being pressed up into the upper and 

 narrower part of the tube; and when it falls, on the contrary, it settles down 

 into the wider part, and becomes shorter, being always the same in quantity. 

 By this means, as the angle of the concave cone of glass, of which this tube 

 consists, is smaller, the different situation of the Mercury will, on the altera- 

 tion of the air's pressure, be nicely shown by very large and distinct divisions. 



Now the use to which I would apply this contrivance of the barometer, is to 



VOL. VI. 3 S 



