OF ARTS AND SCIENCES. 483 



mica plate, through which the samples could be introduced and the 

 ignition noted. 



In order to avoid loss of heat by radiation the bath was surrounded 

 by a loose envelope of asbestos paper, which allowed the hot gases 

 from the lamps beneath to circulate about the bath. For determining 

 temperatures below 200° I used a small mercurial thermometer ; for 

 temperatures above 200°, an air thermometer with constant air ma- 

 nometer, on the plan described some time since by Professor J. P. 

 Cooke,* This thermometer proved to be extremely convenient, and 

 its indications agreed closely with those of the mercurial thermometer 

 in the neighborhood of 200°. 



In preliminary trials, which I made with sound pine wood, I soon 

 found that the ignition point was greatly affected by the way in which 

 the sample was heated. If the temperature was slowly raised, the 

 wood was thoroughly carbonized before the ignition point was reached, 

 and the observed ignition point was merely that of charcoal. Thus 

 four samples of pine-wood, slowly heated, ignited when the thermome- 

 ter reached 405°, 407°, 415°, and 417°, while freshly burned pine 

 charcoal ignited at 403°, 405°, and 408°, at successive trials. Pieces 

 of the same wood (when the bath was quickly heated) ignited at 360°, 

 365°, and 372°, and when dropped into the heated bath they also took 

 fire at lower temperature. 



In order to determine the lowest possible point of ignition, it was 

 therefore evidently necessary to make successive trials, in each case 

 introducing the sample into the bath heated to a definite temperature, 

 and lowering this temperature a few degrees at a time until the sample 

 failed to ignite. Small pieces of wood (0.5 gram) usually ignited, if 

 at all, within fifteen or twenty minutes ; larger i^ieces (15 grams) some- 

 times required heating for an hour or longer. 



The samples of wood which I examined were all portions of timbers 

 from the Calumet and Hecla Mine, and varied greatly in physical 

 properties. They were : — 



* These Proceedings, XVII. 22. 



