78 Mr. Spottiswoode [March 31, 



of rings, each of which would be regarded of ample size for currents 

 obtained in the ordinary way. These are now all connected together 

 so as to form one grand negative terminal ; and it will be found that 

 with the currents from the alternate machine the whole system is 

 readily illuminated at once. 



It should perhaps be here remarked that, while the strength of the 

 secondary currents passing through the tube is partly due directly to 

 the strength of the primary currents from the machine, it is probably 

 also in part due to the rapidity with which the secondary currents 

 follow one another. Owing to the latter circumstance the column of 

 gas maintains a warmer and more conductive condition than would 

 prevail if the interval between the discharge was longer; and in 

 consequence of this a larger portion of the discharges can make its 

 way through than would otherwise be the case. 



Before leaving the instrumental part of my discourse, I desire to 

 bring under your notice a modification of the machine which we have 

 thus far used for producing, by the intervention of the induction coil, 

 currents of high tension. This consists of a machine of the same 

 general construction as the other, but having the armatures wound 

 with a much greater number of convolutions of much finer wire. The 

 result is a machine giving off currents of sufficient tension to effect, 

 by direct action, discharges through vacuum tubes, and even in air. 

 The currents are of course alternate ; but by diminishing the size of 

 one of the terminals to a mere point, as well as by other methods 

 described elsewhere, it is possible to shut off the currents in one 

 direction, leaving only those in the other direction to discharge them- 

 selves through the tube. I ho^De on some future occasion to give a 

 fuller account of this remarkable machine, which has only quite 

 recently been completed. 



Returning to the discharge in air, it will be noticed that when the 

 terminals are set horizontally the torrent of thick discharges assumes 

 the appearance of a flame, which takes the form of an inverted V. 

 This is the result of convection currents due to the heat given off by 

 the discharges themselves. The discharges are by their nature as it 

 were fixed at each end, but within the limits of discharging distance 

 free to move about and to extend themselves in space, especially in 

 their central part. Further, it may be observed that the length of 

 the spark which can be maintained is greater than that over which it 

 will leap in the first instance. The explanation of this is to be sought 

 in the fact that when the sparks follow very rapidly in succession, the 

 whole path of each discharge remains so far in a heated state as to 

 assist the passage of the next ; and, further, that in the middle part 

 of the discharge or apex of the \, where the heat is greatest, the heat 

 prevails to such an extent as to render a portion of the path highly 

 conductive. This may be illustrated by holding a gas jet near the 

 path of the discharge. The flames will then leap to the two ends of 

 the jet, which will perform the part of a conductor; and the real 

 length of the discharge will be that traversed from terminal to 



