190 F. Leininger: Relation of Electric Charges transported 
importance, the values for the canal rays exceed those for 
cathode rays, since then the longer path of the latter is of 
greater importance. 
TABLE LV, 
Canal rays in rarefied air. Second kind of net-electrodes. 
) / 


) | Value of Values corrected 
Potential in volts. | 199 canal rays for absorption 
current by net. 
600 | 9°7—10°2 |. 149 ae 
1200 | -22 —22°8 42-4 —44 
1600 | 23 —24°4 44-4 —47-] 
1800 : 23'8 —24-7 459 —47-7 
2000 24-6—25'8 f 47°75 —49'8 
2200 25°'2—26°5 ) 486 —dl1'1 
- 2400 26°7—27 | 51°55—52°1 
3000 27-295 8 8| 521 —569 | 
4000 28°7—29°9 ) 554 —Sr 
4700 ) 29°9—31°5 ! 57°7 -—60°8 
5000 30°7—31°8 59°2 —61°4 
6000 ) 31°-4—32°6 . 60°6 —62°9 
7000 32°7—34 ) 63:1 —65°6 
Sources of Error. 
A considerable source of error in carrying out the expe- 
riments arose in the gradually developing inconstancy of the 
vacuum; since the current, as mentioned above, in passing 
through it freed the occluded gases from the metal portions 
of the tube. A complete suppression of this evolution of gas 
could not be attained. It was sought to diminish its effect 
on the results by always sending the current through the tube 
fora very short time only while taking the actual observations, 
and for the remainder of the time keeping the influence- 
machine short-circuited. 
With the form of tube shown in fig. 2, this evolution of 
occluded gases could more readily be reduced to a minimum. 
For this reason the tube was used in making the first mea- 
surements, strong discharges being maintained through it by 
means of an induction-coil for several hours while the ex- 
haustion was going on. 
Eaperiments with Pure Gases. 
In order to arrive at some relation connecting the ratio 
of the rays to the current with the nature of the gaseous 
contents, the former experiments were repeated with tubes 
