Doppler Effect in Positive Rays. 521 



The form of tube I employed is represented in fig. 1. 

 The slit of the Rowland grating was at X for the retrograde- 

 rays and at Y for the canal rays, 

 j?i<r, l. A being the anode and C the cathode* 



The Rowland grating: gave, in the 

 order of spectrum I employed, six 

 [T\ Angstrom units to *9 mm. The 



\ / effect was observed with respect to 



the hydrogen line 4861*5, and the 

 change in refrangibility was mea- 

 sured by comparison with the solar 

 spectrum, which was photographed 

 immediatelv beneath the gaseous 

 line without changing the iaws of 

 the slit. 



The amount of the change in re- 

 frangibility was sensibly the same 

 as in the canal rays. 

 The difference of potential between the anode and the 

 cathode varied between 5000 and 10,000 volts ; and the 

 current from ten milliamperes to five, furnished by a storage- 

 battery of 10,000 cells. A current of running water provided 

 a large and steady resistance. 



The appearance of the discharge at the cathode has often 

 been described. The cathode appears to be the base of two 

 rose-coloured cones of light, the apex of one directed to the 

 anode and the apex of the other toward the canal region. 

 The body of the luminous cone in the space between the 

 anode and the cathode is, so to speak, a solid, while that in 

 the canal region, or back of the cathode, is made up of a 

 collection of tubes which in a short region come together at 

 the apex ot the cone, and in a more extended region spread 

 out in a diffused manner. 



When the cathode is imperforated the rosy glow which, in 

 the case of hydrogen, characterizes the canal rays, emanates 

 from the central portion of the aluminium cathode ; it is no 

 longer conical in form, or rather resembles a frustrum of a 

 cone, the base directed to the anode. It does not extend as 

 far toward the anode as the conical discharge from the 

 perforated cathode, and is not so bright. 



The form of tube I employed is represented in fig. 1. 

 The anode was placed in a side tube which was at right 

 angles to the tube containing the cathode C. There was 

 thus the possibility of placing the slit of the spectroscope 

 either at X or at Y. 



When an imperforated cathode was employed and the slit 



