522 PROCEEDINGS OP THE AMERICAN ACADEMY. 



As the electrolyte became hot the intensity of the light became less, 

 and for the sake of economy in time the solution was usually changed 

 every ten or fifteen minutes. The points of most of the metals used 

 also required frequent replacing, and this was done by feeding the wire 

 down through a simple clamp as fast as it was eaten away. 



The first series of observations showed that the spectrum is exactly 

 the same, whether the metal in question is used as active electrode or is 

 present as a salt in solution ; further, whether the metallic point is anode or 

 cathode, though a great difference in the intensity of the light exists in 

 the two cases. The metallic point as anode gives only a feeble zone of 

 light. As cathode several distinct stages of the condition about the 

 point are to be distinguished, one of these being accompanied by a 

 rapid " zerstiiubung " of the metal. This condition is unfavorable for 

 the production of a bright luminescence, but goes over with increase of 

 current density into the more favorable condition. 



It was found that the spectrum is exactly the same whether produced 

 by direct or alternating current. With direct current the electrolyte 

 heated faster and the point was more rapidly eaten away. The greater 

 part of the work was therefore carried out with the commercial alternat- 

 ing current, which is of 110 volts and 60 cycles. The current through 

 the cell averaged 2.5 amperes, and it was found best to keep the resist- 

 ance of the electrolyte at a point which would give about this current. 

 The large plate had a contact surface of about 10 sq. cm. and the wires 

 used were of diameter 0.3 to 1.5 mm. 



No direct experiment on the effect of inductance in the circuit was 

 made, but the cell was used alternately without inductance and as inter- 

 rupter with a large induction coil for the production of the spark spectra, 

 without any change visible in the small direct vision spectroscope. 



It was expected that many gaseous lines would appear, and they were 

 looked for, but the only one found under the conditions of the experi- 

 ment was the red hydrogen line at A 6562.* This is visible clearly 

 in many cases ; in others it is either very faint or entirely invisible. 



In the following tables the wave-lengths of the principal lines are 

 given, with their intensities in the spark, arc, and Wehnelt spectra on a 

 scale of 1 to 100; 1 is the intensity of a barely measurable line, and 

 100 that of very strong lines, like the principal lines of the alkali- and 

 alkaline earth-metals. It is of course clear that such intensity measure- 



* Werner von Bolton used the hj'drogen lines of the Wehnelt as standards for 

 comparison, and shows them in his drawings of spectra. Certainly none appear on 

 any of the author's plates. 



