SHORT NOTES ON PROFESSIONAL SUBJECTS. 
99 
Originally it was proposed to use the spark as the recording agent in the first 
instrument constructed, but it was abandoned in that case because it was believed 
that there was a considerable amount of uncertainty in its action. Even at the pre¬ 
sent day it does not appear that any satisfactory experiments have been made 
by its aid, although it has so frequently been proposed to use it. The velocities 
obtained by the use of Schultz and Vignotti’s instruments when they were placed 
in the same circuit differed remarkably, and in both instruments the records were 
made by a spark from a coil*. Still this would be a most useful recording agent 
if it could be relied upon. An attempt has therefore been made to retain it. 
Between the conducting point and the paper a piece of glass is placed in which there 
is a conical hole having a small opening, and the discharges are made from the point 
f Fig. 2 through this opening. If the spark be not satisfactory then mechanical 
means must be used. 
The following is an arrangement of a much more simple nature, but it would 
require the exercise of care in adjusting the strength of the galvanic 
currents. Suppose that the first screen, an electro-magnet capable of sup¬ 
porting a ball, and an electro-magnet capable of working a relay, are all brought 
into one galvanic circuit. Suppose the same thing done for the second screen 
but quite independently of. the first screen. When the first screen is broken the 
corresponding weight is dropped and the primary current of the Ruhmkorff’s 
coil is broken so that a record a Y is made on the revolving cylinder. When the shot 
breaks the second screen the second ball is dropped and a spark record Zq is made on 
the cylinder. When the ball liberated by the shot at the first screen has fallen 
through a space corresponding to some known time, as half a second, it permanently 
stops the first current and gives a record a 2 on the cylinder. When the second ball 
has fallen through an equal space it gives a record Z> 3 . The records will stand 
thus— 
a \ &i a % > ^2 
where a 1 a 3 represents the time occupied by the first ball in falling through its 
assigned space, and ZqZ> 3 the time occupied by the second ball in falling through an 
equal space, a 1 a 3 ought therefore to be equal to ZqZ> 3 if the angular velocity of the 
cylinder has continued uniform, or aqa 3 ought to be very slightly greater than ZqZ> 3 if 
there has been any slight loss in the angular velocity of the cylinder. aqZq, or a 3 Z> 3 
will represent the time occupied by the shot in passing between the two screens. 
This arrangement is open to the objection that the records may be affected by re¬ 
maining magnetism, just as in the arrangements of Navez, Leurs, or Boulenge, but 
it possesses the recommendation of greater simplicity. 
* Beliefs Electro Ballistic Machines. New York, 1866. p. 47. 
