Observations on the Spark Discharge. 



79 



are there constrained in their motion or are fewer in number. If an 

 external by-path be arranged, the spark will often prefer to clear as 

 much as 5 cm. of free air to passing a distance of 1"5 cm. between the 

 glasses. Accurate measurements were not attempted, for with the 

 apparatus used it would evidently have been difficult to have 

 arranged that definite conditions should obtain as to the space be- 

 tween the glasses. 



On arranging the slip beneath the receiver of an air jmmp so that 

 the path of the spark can be observed at low pressures it is some- 

 times, not always, seen that at low pressures the curvature of the 

 spark increases. It may move outwards by three or more rings, the 

 rings themselves remaining unaltered. When the pressure is much 

 reduced it becomes difficult any longer to keep the spark between 

 the glasses. It chooses then some external path. 



(15.) In the last experiments after the first half-dozen sparks were 

 passed through the straitened space between the glasses a faint 

 etched line becomes apparent at the place where the constraint was 

 greatest, i.e., close to the rings. The formation of this line may be 

 observed. It is accompanied by a local brightening of the spark and 

 an emission of sodium tinted i^ays as if the glass was being volatilized. 

 The line will preserve the curve described by the spark. It may be 

 as long as 4 or 5 mm. or merely a speck. 



On placing this under the microscope the line is resolved into a 

 very beautiful and regular structure, diatom-like in its delicacy of 

 marking. Fig. 1, PI. 3 is a photograph of such a spark track magni- 

 fied 64 diameters. The width of the track is about one-tenth of a 

 millimetre, but this is slightly variable. Measurements of the side 

 rays show that they are from 330 to 420 to the millimetre. Such a 

 track may be deepened and lengthened by repeatedly passing the 

 spark, the side rays remaining apparently undisturbed in position, 

 but widening a little and lengthening a little, so that a track which 

 is a very faint impression when first examined may be submitted 

 again to the current and gradually brought up. It will be found 

 too in this process that the pitted central core of the track widens 

 and sinks deeper into the glass. The contrast between a fresh and a 

 worn track is interesting. In the former none of the central pitting 

 is developed, but the side rays meet at the centre somewhat as shown 

 in the cut (fig. 3). If the glasses be taken asunder it will be 



Fig. 3. 



Formation of spark tracks. 



