110 PHILOSOPHICAL TRANSACTIONS. [aNNO 1786, 



the bottom of the glass ball hermetically, and by 

 that means cut off all communication between 

 the air confined in the ball and the external air ; 

 and with the instrument so prepared I repeated 

 the experiment before-mentioned; that is, I put 

 it into water warmed to 1 8°, and when it had 

 acquired the temperature of the water, I 

 plunged it into boiling water, and observed the 

 times of the ascent of the mercury in the ther- 

 mometer. They were as annexed. 



Hence it appears that the Torricellian vacuum, which affords so ready a pas- 

 sage to the electric fluid, so far from being a good conductor of heat, is a much 

 worse one than common air, which of itself is reckoned among the worst : for 

 in the last experiment, when the bulb of the thermometer was surrounded with 

 air, and the instrument was plunged into boiling water, the mercury rose from 

 18° to 27° in 45 seconds ; but in the former experiment, when it was surrounded 

 by a Torricellian vacuum, it required to remain in the boiling water 1 minute 30 

 seconds = QO seconds, to acquire that degree of heat. In the vacuum it re- 

 quired 5 minutes to rise to 48°-^ ; but in air it rose to that height in 2 minutes 

 40 seconds ; and the proportion of the times in the other observations is nearly 

 the same. 



To remedy some inconveniencies in the instrument, I had recourse to another 

 contrivance much more commodious, and much easier in the execution. At the 

 end of a glass tube or cylinder, 10 or 11 inches in length, and near 4 of an inch 

 in diameter internally, I caused a hollow globe to be blown l-f inch in diameter, 

 with an opening in the bottom corresponding with the bore of the tube, and 

 §qual to it in diameter, leaving to the opening a neck or short tube, about an 

 ijichor 4- of an inch in length. Having a thermometer prepared, whose bulb 

 was just half an inch in diameter, and its freezing point fell at about 2^ inches 

 above its bulb, I graduatefl its tube according to Reaumur's scale, beginning at 

 0°, and marking that point, and also every 10th degree above it to 80°, with 

 threads of fine silk bound round it, which being moistened with lac varnish ad- 

 hered firmly to the tube. This thermometer I introduced into the glass cylinder 

 and globe just described, by the opening in the bottom of the globe, having 

 first choaked the cylinder at about 2 inches from its junction with the globe by 

 heating it, and crowding its sides inwards towards its axis, leaving only an open- 

 ing sufficient to admit the tube of the thermometer. The thermometer being 

 introduced into the cylinder in such a manner that the centre of its bulb coincided 

 with the centre of the globe, I marked a place in the cylinder, about 4 of an 

 inch above the 80th degree or boiling point on the tube of the inclosed ther* 



