86 



this lamp, the current of air impelled by the fans had been 

 made to pass over the naphtha in hundreds of streams before 

 it could arrive at the burner. By allowing a sufficient quan- 

 tity of common air to mix with the naphthalized air, as it 

 passed into the chambers of the burner, a degree of dilution 

 would be produced that would cause the mixture to burn with 

 a pure white light. 



" On constructing the lamp with various valves to effect 

 this object, I found that, unless the common air intended for 

 dilution were allowed to mix with the naphthalized air in se- 

 veral streams, the common air would take a direction through 

 one or other of the gas-ways of the burner, and the resulting 

 flame would be yellow in one part, blue in another, and white 

 in a third. By passing the common air in a number of streams 

 a perfect commixture was effected ; and by means of a lever 

 outside of the lamp I was enabled to increase or lessen the 

 tenuity of the streams, so that the body of the flame at the 

 burner might be rendered less or more dense, or changed to 

 a flame that should show no light, or to a smoky one. Perfect 

 control was thus established. 



" 1 have found by experiments conducted on a very large 

 scale, that one hundred gallons of naphtha, of s. G. 0*846, 

 distilled from Parrot coal tar, are, in burning with common 

 air, equal, in the light produced, to 122 gallons of the best 

 spermaceti oil. The comparison was made with Argand 

 lamps and lamps of my construction, both showing the same 

 diameter of flame. Other experiments made on a small scale, 

 with photometers of different kinds, gave results a little diffe- 

 rent. In one trial 100 measures of naphtha equalled 120 of 

 oil; in a second 128 of oil; in a third 130. Assuming 125 

 to be the number, and taking naphtha at 3^. 6c?. per gallon, 

 and spermaceti oil at double that sum, the oil lights will be 

 two and a half times more expensive than the naphtha lights. 



"I made photometrical experiments on the comparison of 

 my lights with common gas lights, the result of which was, 



