Morrow — Spectra of 3iangane!ie, Lead^ Copper, and Lithium. 271 



The Apps coil used gave a 12-inch spark ; and the condenser consisted of a sheet 

 of tin foil about 15 inclies by 18 inches on either side of a sheet of glass. 

 The drying tubes contained phosphorus pentoxide and caustic potash to 

 absorb any acid vapours which might be liberated, and were connected with 

 a Geryk vacuum-pump. 



During the experiments tlie pressure in the tubes was about three or four 

 millimetres, wliich was observed by having a tube from the connexions set 

 vertical in a small trough of mercury, witli a barometric tube for comparison. 



Sir Walter Hartley has shown that in almost all cases the spectra of the 

 solutions of salts are the same as the corresponding metals : the non-metallic 

 constituents do not affect the spark-spectra ; and Dr. PoUok has proved tliat, 

 with the chlorides in the vacuum-tube spectra precisely the same lines are 

 obtained as with the vapour of the metal with or without the lines of chlorine, 

 and that these lines agree with those of the spark-spectra. Chlorides were 

 used in preference to other salts, because they are easily obtained, and are 

 readily volatilized without decomposition. 



Four new quartz tubes fitted with new platinum electrodes were used. 

 Into the lower bulb of each tube was put a drop of the distilled water used for 

 making the solutions required for tlie experiments. Tlie electrodes were 

 connected, the tube exhausted, and the current passed, the capillary portion 

 being heated all the time during the passage of the current, and a second 

 burner used for warming the lower limb. Photographs were taken of the 

 spectrum, giving five minutes' exposure in each case. To facilitate the 

 identification of the linos, on the centre of each photograph the spark-spectrum 

 of cadmium was superposed, using a Hemsalech coil for the removal of air 

 lines. Tlie exposure in this case was for one minute. On comparing the 

 spectra which are superposed it will be found that unless tlie two sources of 

 liglit be placed in the same position with regard to the slit, and the rays 

 fall on it at the same angle, the photographs of the spectra will be displaced 

 relatively to each other, but if the image of the vacuum-tube be focused on 

 the slit by means of a spherical lens, and this lens be not moved before the 

 cadmium spectrum is photographed, the two spectra on tlie plate will be quite 

 correctly placed in relation to each other. Care should be taken that tiie light 

 source, the slit, and the centre of the collimating lens are in the same straight 

 line as seen by vertical pointers attached to the instrument. Of course in each 

 case the cadmium electrodes and the centre of the capillary portions of the 

 tubes were equidistant from the slit. The lines in the cadmium spectrum 

 were shortened by means of an aluminium sliding sliutter with a horizontal 

 V-shaped notch. 



The photographs obtained from the distilled water showed water-vapour 



2t 2 



