and in the Condensation of Vapours. 389 



often as I made two consecutive experiments without 

 filling the lamp afresh, the alcohol constantly appeared 

 weaker in the second experiment than in the first. 



The cause of this phenomenon was not difficult to dis- 

 cover. The most volatile and consequently the most 

 combustible parts of this liquid, being diffused in vapour 

 in the interior of the lamp, found means of escaping 

 through the nozzle with the part of the liquid that trav- 

 .ersed the match, leaving the alcohol that remained in the 

 lamp perceptibly weakened. 



To remedy this imperfection, I constructed a third 

 lamp, which I now submit to the inspection of the Class. 

 It is made of copper, and has the shape of a small cylin- 

 dricaf vase, an inch and a half in diameter, and three 

 inches high, swelling out a little atop, and closed hermet- 

 ically by a copper stopple, which, being ground with 

 emery, fits tight into the neck of the vase. Through 

 the centre of this stopple passes a small perpendicular 

 hole, which can be shut completely or left a little open, 

 as may be required, by means of a small screw carrying 

 a copper collar. 



A small tube, about an eighth of an inch in diameter 

 and two inches and a half long, proceeds horizontally 

 from the side of the vase very near the bottom. At the 

 distance of an inch and four lines from the vase this tube 

 is bent at a right angle, rising upwards perpendicularly 

 to form the nozzle of the lamp. 



This little tube is everywhere very thin, except at its 

 upper extremity, where it is made thicker, to admit of 

 being shaped so as to fit tight into a very small cylindri- 

 cal extinguisher, five lines high by three and a half in 

 diameter, intended to close the nozzle hermetically with- 

 out touching or deranging the wick, the moment the 



