316 PROCEEDINGS OF THE AMERICAN ACADEMY 



Dr. W. F. Channing had formerly experimented with cam- 

 phene and other chemical burning-fluids, and he was satisfied 

 that they do not spontaneously explode, and that they do not 

 form an explosive mixture with atmospheric air, without the 

 odor of the fluid becoming perceptible to the sense of smell. 



Dr. J. Bigelow remarked, " that the condition of a canister 

 having one of its apertures stopped with a porous body, was 

 like that of a common carnphene lamp with a tube and wick. 

 An explosion would not be likely to be communicated through 

 the porous body, nor would it take place unless some open 

 aperture communicated with an explosive mixture within. He 

 mentioned a remarkable case, which occurred some years ago, 

 in the chemical laboratory of the old Medical College. The 

 iron pipe of a stove, containing a fire, passed within a foot of 

 a shelf on which were deposited some bottles containing dif- 

 ferent volatile oils. In the night the whole took fire, and in 

 the morning the shelves and side of the apartment were 

 found deeply charred, and the room filled with smoke. The 

 fire, however, was spontaneously extinguished. On examina- 

 tion, it was found that a lead pipe, communicating with a 

 water-cistern above, had been melted ofi", and the water had 

 flowed down upon the fire. The bottles which contained 

 the oil were found in their places, some broken, others with 

 their stoppers blown out, with appearances indicating com- 

 bustion rather than violent explosion." 



Dr. C. T. Jackson said " he had listened to the ingenious 

 explanation of Professor Horsford, and would take occasion to 

 remark, that he could not conceive how a spark from pine- 

 bark tan could set fire to the vapor of burning-fluid, even 

 allowing the spark should have passed near the slightly stop- 

 pered can. It is well known that a red-hot coal will not 

 kindle a flame in camphene or burning-fluid vapor, and that 

 actual flame or incandescent heat is required to inflame va- 

 pors of volatile hydro-carbonaceous fluids. 



" If there was no other way to account for the combustion 

 of the vapor from this burning-fluid, he would suggest that a 



