240 



Also, very often when adjusting the speed to obtain photographs for 

 series B and E the discharge would change from one type to the other 

 when the magnets were excited. Preclit found that this was the case, 

 but these observations can hardly be compared with his, since point elec- 

 trodes were nsed in this case, while he used one point and one blunt elec- 

 trode. In all cases observed where a change occurred, if a brush dis- 

 charge in a nonmagnetic field passed above or below a line directly be- 

 tween the points as shown by the spark discharge E, 1 and 4, and the mag- 

 nets were excited to deflect the stream in such a way as to make the path 

 of discharge shorter, it changed to a spark discharge. Or if a spark dis- 

 charge passed directly between the points and was deflected it changed 

 to a brush. In all observed cases (possibly 25 or 30) the transformation 

 could be explained by the change of distance. 



The series G shows the effect of an air current on the path of dis- 

 charge. The air current was led into the camera through the bottom side 

 by means of a glass tube 2.25 cm. in diameter so that the mouth of the 

 glass tube was 2.2 cm. below the points, and flowed at the rate of about 

 1,200 c. o. per second. Nos. I and 3 show the discharge without the air 

 current, and Nos. 2 and 4 show deflection by the air current. It differs 

 from the deflection produced by the magnetic field in that the greater de- 

 flection here is with the negative stream. This indicates that the pressure 

 is not as great in the negative stream as in the positive, which agrees with 

 the work of 1 S. Arrhenius. who measured the torsion produced by a sus- 

 pended wire cross with points bent at right angles to point in the same 

 direction and found that the torsion produced by the negatively charged 

 wire was less thin the positively charged wire, which was more clearly 

 shown the lower the potential. (Note — It was intended to show a photo- 

 graph with current in second direction, deflected by an air current. G 4, 

 which should have shown this, shows a current in the same direction as 

 G 2. which was due to a reversal of polarity of the machine. The error 

 was not observed until the apparatus was torn down.) 



Series H shows photographs of the points when the poles of the ma- 

 chine were placed close enough for a spark to pass between them. It was 

 found that when a spark passed between the poles of the machine there 

 was a violet stream (brush) between the points. This violet stream did 

 not usually pass directly from one point to the other, but was curved with 



S. Arrhenius (Annal. Phys. Chem. G3, pp. 305-313), 1897. 



