332 Messrs. A. Schuster and G. Hemsalech. 



revolutions was generally about 120, giving a linear velocity of about 

 100 metres/second for that part of the film on which the photograph 

 was taken. 



The electric discharges were obtained from a battery of six Leyden 

 jars, having a total capacity of 0*033 microfarad, and being charged 

 from an induction machine constructed for us by Mr. H. C. Wimshurst. 

 This machine has twelve plates of 62 cm. diameter, and gives sparks 

 which are 13 inches long. The electrodes were, as a rule, placed 

 1 cm. apart, and an image of the spark was projected on the slit of 

 the spectroscope, the distance of the slit from the electrodes being 

 equal to four times the focal length of the projecting lens, so that the 

 image was equal in size to the spark. The prism used was made by 

 Steinheil, and had a refracting angle of 60. 



We may now pass to the description of the results obtained when 

 the spectrum of a single spark is taken on a moving film. A pre- 

 liminary trial with various metallic electrodes had shown us that the 

 sharpest results were obtained with zinc, and we therefore chose that 

 metal for our first investigation. The principal lines of zinc as they 

 appear on our photographs are the double line, the least refrangible of 

 the two having a wave-length 4924*8, and the blue triplet, the wave- 

 length of the leading line being 4810*7. All the lines are curved on 

 the photographs taken with the spinning disc, but the displacements, 

 especially near the poles, are subject to considerable variations. This 

 is probably due to the fact that the path of the metallic particles is 

 not always straight, and, if straight, its image does not necessarily 

 coincide with the slit. A very slight error in measurement will also 

 affect the results considerably when the total displacement measured is 

 small. Our results do not for this reason allow us at present to give 

 any opinion as to the maximum velocity of the particles near the pole ; 

 but if these are considerable, they drop down very quickly to speeds 

 which, in the case of zinc, are not far off 500 metres/second. 



We have adopted two methods of comparison between different 

 photographs. We have in the first place measured the displacements 

 at a number of nearly equidistant points, and from these measure- 

 ments we have deduced the time taken for a metallic molecule to pass 

 from the pole to a point 2 mm. away from it. If this method could be 

 applied in every case, it would form a rational and consistent basis of 

 comparison. But the curved lines which are to be measured are often 

 very diffuse near the pole ; this, and the continuous spectrum, may 

 render it impossible to obtain satisfactory measurements at that point. 

 In order not to have to reject unnecessarily a large number of measure- 

 ments because the spectrum near the pole was indistinct, we have 

 adopted another method, which, though less rational than the first, is 

 found to give consistent results. From all our measurements we may 

 deduce certain figures for the molecular velocities at different and 



