418 Messrs. J. Frith and C. Eodgers on the 



instructive than those in PI. IV., as the character of the arc 

 alters more with change of voltage than with change of 

 current. As before, the curves for the solid carbons are all 

 very close together. In this case they all show a minimum 

 (maximum negative) value at about 55 volts. With com- 

 binations having a cored positive, this minimum becomes 

 more strongly marked and occurs at a lower voltage. 



It has been our practice to examine the image of the arc 

 on the screen and note when any change in the appearance 

 of the arc takes place. It was first noticed by Mrs. Ayrton 

 that with cored carbons under certain conditions a dark space 

 appeared somewhere near the centre of the arc, dividing the 

 purple glow into two parts. This was never noticed in the 

 case of solid carbons. Our observations have fully confirmed 

 this, and, moreover, we find that the dark space, when only 

 one carbon is cored, is nearer this carbon. We have noted 

 the appearance or disappearance of the dark space, and we 

 find a remarkable coincidence between this point and the 

 minimum of the curves. In the curves for cored carbons in 

 PL Y., to the right of the minimum point the arc always 

 shows a dark space which becomes less marked on approach- 

 ing the minimum, and at the minimum finally disappears. To 

 the left of this point, although the dark space has quite dis- 

 appeared, yet the difference in colour between the two parts 

 of the glow is still observable, that near the negative carbon 

 being of a redder tint than that near the positive. 



PLY. fig. 2 shows the relation between the length of the arc 

 and its resistance at constant current. The general charac- 

 teristics of these curves are the same as those of fig. 1, since 

 the length of the arc is roughly proportional to the voltage. 



Effect of the Core, 



Since the presence of a core in either carbon has such a 

 marked influence on the resistance of the arc, we tried the 

 effect of boring out the core of a Carre carbon. The result of 

 using this hollow carbon as the positive is shown by curve F 

 in PL V. It will be seen that this curve follows very closely 

 that obtained with both carbons solid. The effect was now 

 tried of filling this hollow carbon first with plaster of Paris, 

 and then with kaolin. In both cases the resistance of the 

 arc so produced was positive. Another effect of these cores 

 was to cause the arc to burn at a much lower voltage for 

 a given current and arc length. The arc was, however, too 

 unsteady to allow of a curve being taken. 



