column and other phenomena connected with the discharge. 309 



might be adduced to explain the appearance. If that be the 

 case it is remarkable that the doubleness should be destroyed 

 by altering either the current or the pressure, and that above 

 and below certain limits of pressure it cannot be produced in 

 a given tube by the intentional introduction of an impurity. 

 The only effect of adding impurity is to cause what were originally 

 well-defined striae to become hazy. 



In order to make certain that it was not due to gas given 

 out by the electrodes, a tube was filled with hydrogen after the 

 electrodes had previously been made very hot several times by 

 running, and before a discharge was passed the pressure was 

 reduced to an amount suitable for giving double striae. When 

 the current was passed they immediately appeared, no matter 

 in what way the hydrogen had been prepared, but by letting in a 

 further amount of gas they were immediately destroyed. 



An attempt was made to see if the different parts differed 

 spectroscopically. For this purpose the vertical discharge tube 

 was covered with black paper and a narrow horizontal slit made 

 at one position, so that by bringing this opposite any striation 

 the latter could be studied by means of a direct vision spectro- 

 scope which was held a centimetre away. In order to intercept 

 light reflected from other parts of the tube, another piece of black 

 paper with a small slit in it was placed between the tube and 

 the spectroscope. As far as could be seen by this method both 

 parts of a double striation gave the same spectral lines, those 

 seen in the spectroscope being the red, green and blue of hydrogen 

 and the green line of mercury. The only difference between the 

 two was that the blue line was stronger in the half nearest the 

 kathode than in the other, while for the red the reverse was true. 

 In the case examined the halves of the striations were separated 

 from each other by about 2 mm., in the photograph of those in 

 moist air the distance between the two is about half this. In 

 the latter case also a change in pressure of less than 1 mm. was 

 sufficient to cause the double striae to disappear when well-defined 

 single striae made their appearance. 



Conditions influencing the distance between the striae. 



6. The effect of altering the current on the distance apart 

 of the striae when other conditions are kept constant was first 

 examined; one division deflexion of the galvanometer corre- 

 sponds to 2-5 x 10~ 5 amperes. In taking readings the current was 

 only kept on long enough to enable the measurements to be 

 taken as otherwise it was found that the distance apart altered, 

 due probably to the heating effects of the discharge and consequent 

 change in density. 



22—2 



