304 Mr Willows, On the distance between the striae in the positive 



accurately in this manner, and, in addition, it could not be varied 

 in wide enough limits ; as the method was unsatisfactory for other 

 reasons it was abandoned as far as the quantitative part of the 

 experiments went, and a battery of over seven hundred storage 

 cells used instead. This entailed a shortening of the tubes as 

 the available voltage was reduced, and this meant, of course, that 

 a smaller number of striae could be obtained, making the accurate 

 measurement of their mean distance less certain, but this was 

 more than compensated for by their increased steadiness. In 

 addition the current through the tube and the difference of 

 potential at the terminals could now be easily measured, the 

 former by a galvanometer in the main circuit and the latter by an 

 electrostatic voltmeter. 



As observed by Hittorf 1 a less voltage is required to maintain 

 the discharge than to start it ; in most cases it was started by 

 wrapping the wires from the secondary of a small coil round the 

 glass in the neighbourhood of the terminals. 



Although quantitative results could not be obtained when the 

 discharge was produced by a coil, yet some interesting qualitative 

 results were noticed ; these are given later. 



3. The gases concerning which definite information can be 

 obtained are very limited, for it is well known that most compound 

 gases are decomposed wholly or partially by the passage of the 

 discharge' 2 ; hence, although the distance between the striae were 

 measured, the proportion of decomposed gas present would not 

 be known, as it would in all probability depend on the pressure, 

 current, &c. and therefore no accurate knowledge would be 

 gained, especially as Crookes 3 has found that when the discharge 

 passes through a mixture of gases each gas gives rise to a separate 

 set of striations. This instability under the electric discharge at 

 once excludes those vapours which give rise to the finest striae, 

 viz. the numerous organic vapours, and also the more common 

 compound gases. Of the simple gases remaining chlorine is 

 difficult to work with as it attacks both electrodes and the 

 mercury in the pump, and oxygen is partially converted into 

 ozone. If ozone were not produced, the latter gas would be very 

 troublesome owing to the extreme feebleness of the light of the 

 discharge, making it impossible to measure the distance between 

 the striae with any approach to accuracy. 



The gases finally used were air, hydrogen and nitrogen. The 

 hydrogen was generally prepared electrolytically, the hydrogen 

 peroxide produced can be taken out by passing through solution 



1 Wied. Annul. 20, p. 705. 1883. 



2 Recent Researches, pp. 178 — 9. 



3 Presidential Address to the Soc. of Telegraph Engineers, 1891. 



