64 Scientific Intelligence. 



the spectrograph directly. It was easy to make corrections for 

 the small amount of light from the discharge which reached the 

 window by sundry reflections from the inner walls of the glass 

 tubes. 



The vapors from volatile liquids were admitted to the jacket 

 tube at the same place where gases were caused to enter from 

 reservoirs. When it was desired to work with the vapor of a 

 volatile solid, fragments were simply placed in the jacket tube, 

 the window being temporarily removed to admit them. To 

 observe the temperature of the various flames, a similar appara- 

 tus was employed, but with the jacket tube sufficiently prolonged 

 to contain a mercury thermometer, which was placed so that its 

 bulb was enveloped in the flame. A small, single-prism, quartz 

 spectrograph was used, and good negatives of the spectra of the 

 flames were usually obtained with an hour's exposure or less. 



The following summary of results may be of interest : Sul- 

 phureted hydrogen mixed with ozone gave a sky-blue flame. The 

 temperature recorded was 70° C. However, this value depends 

 on the pressure. The spectrum was continuous from A 2300 to 

 A 4800 with superposed bands from A 2300 to about A 2600. In 

 this case, as in all the others, the glow ceased at the instant when 

 the electric discharge was stopped. The combustion only occurs 

 at these low temperatures and pressures in ozone; it is not main- 

 tained when oxygen is substituted. Identical results were obtained 

 when a lump of solid sulphur was introduced into the jacket tube. 

 On the other hand, the band spectrum is entirely different from 

 that observed when sulphur burns in air at a pressure of about 

 one atmosphere. Carbon disulphide gave a blue flame, appar- 

 ently the same as sulphur or hydrogen sulphide. The spectrum 

 was not examined. Sulphur dioxide gave no glow with ozone. 

 Only when selenium was gently heated did it give a glow, and 

 this was very inconspicuous. Arsenic was apparently not acted 

 on by ozone. Iodine gave an orange glow, redder than that due 

 to nitric oxide. It is oxidized to a yellowish white solid, proba- 

 bly iodine pentoxide. Ammonia, carbon monoxide, hydrogen, 

 nitrogen, nitrous oxide, methane, ethylene, the vapors of petrol, 

 ether, and alcohol gave no glow. Acetylene, in contradistinction 

 to most hydrocarbons, gave a fairly conspicuous, bluish-green 

 glow. The thermometer indicated 100° C, under the particular 

 conditions of the experiment. The spectrum consisted of the 

 Swan bands, together with the hydrocarbon band at A 4315, and 

 it is thus identical with the spectrum of the inner cone of a 

 Bunsen flame. The color of the flame can be regulated by the 

 acetylene supjfly. With a small supply it is green, the Swan 

 bands predominating. With an increased supply the flame con- 

 tracts, the ozone being now consumed before it has got far from 

 the orifice where it issues into the acetylene. At the same time 

 the color becomes bluer, and the hydrocarbon band brighter, rela- 

 tively to the Swan bands, than in a Bunsen flame. Carbon mon- 

 oxide seems to be the chief product of combustion. Benzene 



