590 Mr. A. M. Tyndall on the 



along the path of the discharge may be uniform enough to 

 cause the sudden current to take the spark form. 



Some interesting results were obtained when N was made 

 larger than P. P was a positively charged point O'Ollo cm. 

 in radius. N was 0*039 cm. in radius, and was backed by a 

 plate which was adjusted at such a distance that though N 

 was larger than P, it started to glow either before or after P 

 according to the distance y between them. In Curves 3 the 

 full line shows the glow-fields at P plotted with y under these 

 conditions. The dotted curve was obtained with N a piece 

 of the finest platinum wire, and hence always first to start 

 discharging. In the region BG of the full curve N started 

 glowing first; near A, P started first, and between A and B 

 there was a transition stage in which, as far as could be 

 ascertained by eye, the two points started together. Curves 4a 

 give a few of the field-current curves obtained with the 

 large N point at varying values of y. The circle on each of 

 the upper curves gives the current at which, as far as one 

 could tell by eye, the point N started to glow. It is probable 

 that the currents thus circled are too great because N could 

 not be viewed in darkness owdng to the proximity of a 

 glowing P. 



It will be seen that in the BC region the current-field 

 curves are similar to those obtained in previous work with a 

 fine N point (Phil. Mag. xx. 1910). In the region AB, 

 where P either started first or simultaneously with N, it will 

 be seen that there was no sudden drop in field at P when N 

 started. It would appear, therefore, that though N was 

 glowing, the negative ions supplied by it to P were not 

 numerous enough to bring the field at P down to that in 

 which negative ions will ionize. As the distance y was in- 

 creased, the proportion of current supplied by N increased 

 and the field at P decreased until the conditions were re- 

 versed and N supplied practically all the current. The 

 field at P was then that in which negative ions will 

 ionize. 



These experiments also throw light upon the effect which 

 was obtained when N was large but not so large as P. It 

 was found in this case that at very short distances the 

 field at P tended to become constant, as is shown in the 

 dotted line of Curves 1 discussed above. Now t at these dis- 

 tances, which were only slightly greater than those at which 

 sparks occurred, N seemed to discharge almost simultaneously 

 with P. Consequently the number of ions supplied to P 

 was small, and hence the field at P was higher that that 

 when N was a fine point discharging vigorously. 



