82 



Mr. J. Joly. 



metre. Bat again measurements in other places, where the spacing 

 was finer, gave similar numbers to those obtained in the case of air, 

 rising to 420 to the millimetre. 



It seems probable, in the first place, that some air will linger in 

 these narrow spaces. This might explain the occurrence of the finer 

 spacing. However, granting this, it is not certain that a wider 

 spacing of the lines in the case of hydrogen may not be explained on 

 its superior conductivity or inferior density. These qualities might 

 very conceivably influence the intensity and effects of the outrush 

 from the spark axis. The spacing of the lines might, in fact, be con- 

 sidered as dependent on the accumulation of pressure at points along 

 the axis of the spark. When this accumulation at any point becomes 

 sufficiently intense it breaks out, cutting a channel. It might easily 

 be supposed that such points, where the conditions were uniform 

 along the path of the spark, would be very evenly spaced. The dis- 

 tance separating such points might depend on many properties of the 

 gas, as its conductivity, specific heat, density, and pressure. There is 

 not then, I think, sufficient reason to suppose that there is anything 

 in common between these marks and the strise observed in vacuum- 

 tubes. The explanation just given seems quite adequate to explain 

 the formation of the marks, but no mere modification of such an ex- 

 planation will fit in with many of the observations on strise. The 

 conditions obtaining are really quite different in the two cases. 



(19.) Whether the period of electric oscillation of the spark path 

 and leads had an influence in determining the spacing of these lines, 

 was investigated by an experiment on a slide in which the dimensions 

 had been altered till nearly double the ordinary size. The tracks 

 obtained in this, however, presented no peculiarities. 



(20.) Observations on the appearance of the spark, confined between 

 the glasses, in the field of the microscope, using a magnification of 

 about 60 diameters, revealed no visible peculiarity. Nor did the use 

 of a wedge of neutral tinted glass, introduced to lessen the dazzling 

 light, enable any detail to be grasped. When, however, passing in 

 a narrow space, the sparks could be seen rapidly melting the glass. 

 Photography of the sparks thus magnified was also tried. The 

 photographs show a bright central part, hazy border, and — very 

 faintly — flame-like streamers extending rectangularly from the 

 border and to some two or three diameters of the sparks at either 

 side. 



(21.) The following experiment is explained in an outrush of 

 matter uniformly from all points along the spark. Into a piece of 

 thermometer tubing about 0*2 mm. in bore and about 4 cm. in length, 

 two leads of very fine, straight platinum wire are laid loosely. They 

 are separated in the tube by about 1 cm. Let this gap be situated at 

 a distance of 1 cm from one end, and therefore 2 cm. from the other, 



