288 Biographical Account of DR WILSON, 



near to one another. At one of the trials, it performed so well 

 as not to vary more than a second in about forty hours, when 

 compared with a very exact astronomical clock near to which it 

 was placed. It was some view of rendering much more simple 

 and cheap the machinery of ordinary movements, by the slow vi- 

 brations of such a pendulum, which induced Mr WILSON to pro- 

 secute these experiments. 



Not long after this, he also put in execution a remarkable 

 improvement of the thermometer, which consists in having the 

 capillary bore drawn very much of an elliptical form, instead of 

 being round. By this means the thread of quicksilver upon the 

 scale presents itself broad, and much more visible than it does in 

 a cylindrical bore of the same capacity; The difficulty of con- 

 structing thermometers of this kind, had nearly hindered him 

 from completing his invention, as the thread of quicksilver was 

 found extremely liable to disunite when descending suddenly in 

 so strait a channel. But, by his long experience, joined to far- 

 ther investigation, and more trials, he at last discovered a method 

 of blowing and filling thermometers with flattened bores, which 

 freed them entirely from this defect. 



About the same time, also, he conceived the design of con- 

 verting a thermometer graduated for the heat of boiling-water, 

 into a Marine Barometer, in consequence of the well-known dif- 

 ference of temperature which water, when boiling, acquires un- 

 der the variable pressure of the atmosphere. This he effected, 

 by making a boiling-water thermometer, about a foot in length, 

 with a pretty large ball, and having a thread of quicksilver as 

 broad and visible as was consistent with a very perceptible run 

 Aipon small alterations of temperature. The stem of this ther- 

 mometer he fortified, by inclosing it in a cylindrical case of white 

 iron, having soldered to it, at its lower end, a socket of brass for 

 receiving half of the ball, which afterwards became entirely de- 



