concerning the Gases of the Atmosphere. 



103 



Experiments upon the visibility of C in vacuum-tubes have 

 shown a much smaller degree of sensibility. The tube was 

 in connexion with a Toepler pump and was traversed by a 

 stream of air. The passage from hi«:h (atmospheric) to 

 low pressure took place at a glass capillary which allowed 

 about 30 c.c. per hour (reckoned at atmospheric pressure) to 

 leak past. When moist air from the room on a damp day 

 (15° C.) was admitted, the hydrogen C-line was very bright, 

 nearly obliterating one of the dark bands of nitrogen. On 

 drying the air with phosphoric anhydride, the C-line dis- 

 appeared. Air mixed with 1 per cent, of hydrogen showed 

 it doubtfully, 1-J per cent, plainly, 2 per cent, perhaps equally 

 with the moist air. The much smaller sensibility (about 

 50 times) in these experiments may be partly due to the less 

 favourable character of the ground upon which the hydrogen 

 line has to show itself. 



Demonstration at Atmospheric Pressure of Argon from very 

 small quantities of Air. 



Success in reducing the necessary amount of air depends a 

 good deal upon the form of tube employed. That sketched 

 (fig. 1) allows a minimum residue to be sparked and examined. 

 In some experiments the tube, standing 

 over weak alkali, was charged with 

 5 c.c. only of air. The first part of 

 the sparking is with electrodes ending 

 in platinum points and brought up in 

 U-shaped tubes of which the bends are 

 filled with mercury. A Ruhmkorff 

 actuated by two or three Grove cells 

 is employed at this stage, and the sparks 

 pass just under the shoulder of the con- 

 taining tube, oxygen being supplied 

 as required from a small electrolytic 

 generator. When the volume is suffi- 

 ciently reduced and most of the nitro- 

 gen has disappeared, the electrodes 

 above spoken of are removed. 



In the next stage the sparks are 

 taken between a sealed-in electrode at 

 the top of the containing tube, and 

 another sealed into the top of a single 

 U-tube brought round through the 

 alkali, and rising (as shown) through 

 the narrow part of the containing tube. 

 In order to avoid splashing and consequent risk of fracture 

 from sudden cooling of the heated glass, it was thought ^n 



Fie- 1. 



Scale 



