of vieiv of the Kinetic Theory of Matter. 



33 



surrounding the heated body into unit heat-flow tubes. The 

 central line or axis of each of these tubes, defined as above, 

 may be taken as unit heat-flow lines. These heat-flow lines 

 from a spherical cup, drawn as just indicated, are shown in 

 fig. 7. 



The appearance of these heat-flow lines at once reminds us 

 of the form of a discharge stream from a spherical cathode, 

 even in a narrow tube. In order to compare the form of a 

 discharge stream from a spherical electrode, with the heat-flow 

 lines from such a body, conditions, similar to those under 

 which the flow lines illustrated in fig. 7 are supposed to exist, 

 must be chosen. For this purpose, the large bulb illustrated 

 in fig. 8 was constructed. This bulb is, as nearly as may be, a 

 sphere of about 15 cm diameter. At the 8 



center of the spherical bulb is placed a 

 spherical aluminum electrode of 2*5 cm 

 radius of curvature, and l'85 cm in diam- 

 eter. The electrode is connected by a 

 platinum wire to the outside of the bulb. 



After exhausting the bulb to a com- 

 paratively low pressure, the spherical 

 electrode is connected to one terminal of 

 a spark gap in the secondary circuit of 

 an induction coil. The whole surface of 

 the electrode appears to take part in the 

 action of discharge, and, at pressures not 

 too low, no very well-defined stream can 

 be distinguished. The appearance of the blue 

 roughly sketched about the electrode in fig. 8. 

 portions of the discharge, from the concave side of the elec- 



discharge is 

 While those 



trode in the direction of the axis of the 

 more dense, yet there is no indication of a 



electrode, appear 

 " focusing" of the 



discharge at the center of curvature of the electrode. As the 

 pressure is progressively decreased, the general blue discharge 

 from the convex side of the electrode, as well as from the 

 edges and outside portions of the concave side, becomes fainter 

 in comparison with the discharge from the central portion of 

 the concave side of the electrode. The glass walls facing the 

 convex side of the electrode appear to fluoresce almost 

 uniformly ; the fluorescence of the walls facing the concave 

 side, falls off very abruptly from the point of intersection of 

 the axis of the electrode with the walls of the bulb. The 

 fluorescence of the bulb on the concave side, at very low pres- 

 sure, appears as a bright spot (not very sharply defined), at the 

 intersection of the axis of the electrode with the walls. At 



A.M. Jour. Sci.— Fourth Series, Yol. VI, No. 31.— July, 1898. 

 3 



