78 
MESSRS. W. DE LA RUE AND H. W. MULLER ON THE 
point was placed under a microscope, and drawn by means of the camera lucida, 
a scale of y^tlis of an inch being subsequently drawn by means of the camera with 
the same magnifying power, and used for laying down co-ordinates for studying the 
shape of the point. 
Fig. 12 represents some of these drawings reduced to half size, the spaces between 
the parallel co-ordinates being Youths of an inch ; those figures without co-ordinates 
are drawn to the same scale. 
The lengths of spark obtained with these terminals positive and a disc are as 
follow : — 
Date. 
No. of 
point. 
No. of 
cells. 
Length oi 
spark. 
inch. 
1876. March 18 
1 
5640 
0173 
„ 
17 
2 
0-200 
„ 
17 
3 
0-243 
18 
4 
0-125 
Diam. 
,, 
21 
5 
„ 
0-230 
inch. 
,, 
22 
6 
0-162 
Platinum wire 
0-009 
,, 
22 
7 
,, 
0-215 
„ 
0-0125 
„ 
22 
8 
,, 
0-198 
Aluminium wire 
0-015 
fused by current previous to 
taking length of spark. 
„ 
22 
9 
,, 
0-182 
Platinum wire 
0-0125 
„ 
22 
10 
„ 
0-2105 
Copper wire 
0-0155 
,, 
23 
15 
,, 
0-1615 
Platinum wire 
0-0125 
fused by current previous to 
taking length of spark. 
April 
11 
21 
„ 
0-229 
,, 
16 
22 
„ 
0-216 
Point accidentally flattened on ascertaining zero point. 
„ 
20 
24 
,, 
0-244 
Oct. 
19 
25 
8040 
0-345 
1877. Eeb. 
13 
26 
,, 
0-342 
Experiments with a point, positive, and negative hollow terminals of different forms 
gave no very marked result, except that a hollow paraboloid gave the longest spark. 
Date. 
Negative terminal. 
Positive terminal. 
Number of 
cells. 
Length of 
spark. 
inch. 
1876. April 10 
Disc IT diam 
Point like 21 
5640 
0-239 
„ 11 
Hollow sphere 1"375 inch diam. 
„ 
„ 
0-220 
„ 11 
Hemisphere 1"375 diam. . . . 
„ 
„ 
0-245 
„ 11 
Hemisphere 0"95 inch diam. 
,, 
„ 
0-230 
„ 11 
Hollow paraboloid 
„ 
>5 
0-255 
With the data obtained by means of the microscope, it was ascertained that 
the longest spark was procured when the point assumed a form resembling a 
paraboloid (points 3, 5, 21, 24, 25, and 26) ; the curved outline, which corresponded 
to that found experimentally, was one in which each succeeding ordinate was in the 
