570 Messrs. Petavel and Hutton on tlie 



a glass window through which the arc can be observed or 

 the spectrum photographed*. In the course of these ex- 

 periments both direct and alternating currents were used, 

 the maximum electromotive force being about 500 volts. 

 With these voltages it was not found possible at the highest 

 pressure to maintain an unbroken arc of sufficient length for 

 satisfactory results to be obtained, and most of the photo- 

 graphs record the spectra of numerous successive flashes. 

 Another apparatus considerably smaller but constructed to 

 work with very much higher voltages has just been com- 

 pleted, and doubtless will prove more suitable for the 

 requirements of this work. 



The spectroscope employed was provided with lenses and a 

 compound Rutherford prism of specially light flint glass, the 

 aperture of the lenses being about 5 cms., and the focal 

 length of the collimator 40 cms. The spectra were photo- 

 graphed on isochromatic plates 15 cms. long, the portion of 

 the spectrum studied being between wave-lengths 5500-3500. 

 The light from the arc was reflected by a mirror, an image 

 being formed on the slit of the spectroscope by means of a 

 lens. Upon removing the mirror the image of a spark 

 placed in line with the axis of the collimator, and at the same 

 distance from the slit as the arc, was projected on to the slit 

 by the same lens. 



In all cases a reference spectrum was taken, this being a 

 spark of the substance iuA'cstigated with self-induction and 

 capacity in the circuit, as fully described by Hemsalechf. 



Carhon. 



A number of photographs of the spectrum of the carbon 

 arc in air at 7, 40, 75, and 100 atmospheres, as also in a 

 mixture of air and hydrogen at 16 atmospheres pressure, 

 were taken. Owing to the rapid burning away of the carbons 

 in air under these pressures the gaseous atmosphere always 

 contained a considerable percentage of carbon monoxide, and 

 in addition, generally, carbonic acid gas. The red vapours 

 of oxides of nitrogen, referred to by Wilson and Fitz Gerald J, 

 were conspicuous by their absence, and chemical analysis 

 failed to show any noticeable formation of nitric acid. As 

 far as could be judged the atmosphere remained fairly clear, 



* A view of the apparatus will be found in Journ. Instit. Electr. EDg. 

 vol. xxxii. p. 236 (1903). 



t Gr. A. Hemsalecli, Comptes Rend. vol. cxxix. p. 285 (1899) ; Journ. 

 de Ph/siqne, (3) vol. viii. p. 642 (1899). 



I Proc. Roy. Soc. vol. Ix. p. 378 (1896). 



