1 24 Management of Light in Illumination. 



goes on, a portion of atmospheric air will make its way 

 into the reservoir by this opening ; and the more the 

 cold increases, the greater will be the quantity of air 

 which will thus find its way into the reservoir. 



As long as the cold continues, this air will produce 

 no bad effects ; but as soon as the lamp becomes 

 sensibly warmed, either in consequence of its being 

 lighted or of a change of temperature in the surround- 

 ing atmosphere, the elasticity of the air confined in the 

 upper part of the reservoir will be increased, and will 

 cause a part of the oil below it to be driven out of the 

 reservoir, which will overflow the burner and run out 

 of the lamp. Various attempts have been made to 

 remedy this capital defect of lamps with inverted reser- 

 voirs, but none of them have been completely success- 

 ful. None of them that I have been acquainted with 

 have rendered it possible to light one of these lamps 

 a second time (without emptying and filling it anew), 

 without danger of having some of the oil forced out of 

 the lamp by the expansion of the air in the reservoir, 

 on its being warmed. 



This accident is always very disagreeable;* and I 

 took special care to avoid it in my illuminators, by 

 avoiding the use of inverted reservoirs. 



As every new contrivance, however simple it may be 

 in its construction, is in the s^reatest dans^er of beinsf 

 put out of order and spoiled by the ignorance and 

 awkwardness of those into whose hands it comes, it is 

 indispensably necessary that the most particular prac- 

 tical directions should accompany every proposal for the 

 introduction of new improvements. On that ground 

 I hope to be excused for giving the following very par- 

 ticular directions for the management of my illuminators. 



