82 Mr. Spoitiswoode [March 31, 



receives strong support from the phenomena represented in Figs. 

 5, 6, and 7. The successive bright lines there shown must be due to 

 successive falls and revivals of tension within a single coil discharge. 

 The existence of such alternations in coil discharges of large 

 quantity is otherwise known. "When the fall in temperature is such 

 that the conductivity of the gas is insufficient to maintain the arc, 

 the discharge can make its way through the air only by a fresh rent 

 of the same kind as the first fracture. But how can this be recon- 

 ciled with the fact that the tension can never reach its original degree, 

 and must, on the whole, be gradually falling, and that, in addition, 

 the paths represented by these various sparks are successively longer 

 and longer? The answer to this question is to be found plainly 

 written in the phenomena themselves. Any irregularity in one of 

 these bright lines is always found to be accurately repeated in all of 

 the same series. Now, it is scarcely to be conceived that, at succes- 

 sive instants of time and in different portions of space, irregularities 

 in the discharge itself, and in the distribution of the gas, so precisely 

 the same, would constantly and for certain recur ; and we are there- 

 fore driven to the conclusion that it is the same portion of gas which 

 at first occupied the centre of the field, with its same yet unhealed 

 rent, which is moved outward under the action of the magnet. If this 

 be so, we have in this repetition of minute details nothing more than 

 what would necessarily follow from successive reopenings of the 

 weak parts of the gas, which would be surely found out by the elec- 

 tricity in its struggle to pass. 



The view here taken of the material character of the luminous 

 discharge is further borne out by the fact that the spindle of light is 

 capable of being diverted by a blast of air. When the blast is gentle, 

 the discharge becomes curvilinear, approximately semicircular, and 

 the yellow flame may be seen playing about the outer edge, in the 

 same way as in a weak magnetic field. When the blast is stronger, 

 the sheet of light becomes irregular in form, and it is traversed by a 

 series of bright lines, all of which follow, even in their minute details, 

 the configuration of the sheet. The analogy between this and the 

 phenomena produced in a strong magnetic field needs no further 

 remark. If the strength of the blast be still further increased, the 

 flame and the sheet of light both disappear, and nothing remains 

 but bright sparks passing directly, and undisturbed, between the 

 terminals. In this case the air is both displaced and cooled so 

 rapidly by the blast, that it no longer offers a practicable conduc- 

 tive path for the remainder of the electricity, coming from the coil, 

 to follow. Of this a succession of disruptive sparks is a necessary 

 consequence. 



The effect thus produced by a very strong blast is in fact similar 

 to that observed when a jar is used as a secondary condenser. In 

 this case the electricity, instead of flowing gradually from the coil, 

 passes in one or more instantaneous discharges with finite intervals 

 of time between them. Each of these has to break its way through 



