310 Mr. G. A. Hemsalech on the Luminous Vapours 

 graphite a, b, c, 



passage towards 



d, care being taken to leave open a free 

 the opposite end so as to enable the ionized 

 vapours to creep along the surface of the plate under the 

 impulse of the electric field due to the heating- current. 

 As before, the hottest part of the plate — namely, the middle, 

 was focussed upon the collimator slit. Observations were 

 made of the vapour on both sides of the plate. 



The spectrum given by copper vapour under these con- 

 ditions is composed of lines and bands ; the latter are par- 

 ticularly well developed above the plate, and are probably 

 due to oxidation of the metal, which is facilitated by the 

 inrush of fresh air through convection. The band spectrum 

 is practically identical with that observed for the same 

 metal in the tube-furnace. With regard to the line spectrum 

 the so-called flame lines are more intense above the plate 

 than in the protected space beneath, but the electrically 

 controlled lines have the same strength in the lower and 

 upper red fringe, although in the latter they pass to a 

 greater distance from the incandescent plate for reasons 

 already explained in § 2. 



The tabulated results refer to the emission of copper vapour 

 in the protected space beneath the plate, and are founded 

 upon a photographic record obtained in the manner described 

 at a plate temperature of about 2900° C. The value of the 

 actino- electric field in the vicinity of the plate was 7*5 15— b . 



The relative intensities of the copper lines in the air-coalgas 

 flame, as observed by Dr. de Watteville *, and in the capacity 

 spark as found by me t have been added for comparison. 





Singi 



e plate furnace. 







Luminous 







Fringe 



Vapour 





A. 



(thermo- 



( therm o- 





electrical 



chemical 



Type. 





excitation). 



excitation 

 at 2800° 0.). 





5105-58 



6 



6 



long 



5153-26 



2 



— 



short 



5218-20 



3 



* — 



short 



5700-25 



1 



1 



long 



5782-16 



3 



3 



long 



Air-coal gas flame. 



Cone 

 (chemical 

 excitation). 



10 

 3 

 5 

 3 



10 



Mantle 



(thermo- 



chemicaL 



excitation 



at 1850° C). 



10 



9 



10 



Capacity 



Spark 

 (electrical 

 excitation). 



10 

 15 



Dr. de Watteville has also very kindly supplied me with 

 copies of his original photographs of the flame spectrum of 

 copper, which go to show that the two lines 5106 and 5782 

 are equally intense in cone and mantle. With regard to the 



* De Watteville, Theses, p. 31. 

 t Hemsalech, Theses, p. 107. 



