IOWA ACADEMY OP SCIENCE 
187 
metal and exert very little moment of force on the key, because of the very 
small time required to pass through. However, we did succeed in making 
keys of very light spring wire which operated quite satisfactorily. A small 
bristle board card was attached to the wire. The whole moving part weighed 
only a little over a gram. The contact was made or broken in a mercury cup 
by a motion of less than one millimeter. The apparatus as used for measuring 
the velocity of rifle bullets is shown in Pig. 4. The rifle is at the left, seen 
only in part. The two keys are just above and back of the galvanometer. The 
bullet hole in the paper screen and the wires leading to the keys may be seen. 
Two sets of observations were taken with this apparatus for Peters 22 short 
cartridges when shot from a rifle such as is used in target practice. The 
data is given in the following table. 1 division corresponds to 0.00054 seconds, 
distance between keys deflection time velocity 
9.8, 
9.4, 7.0, 4.5, 
7.0, 
6.5, 6.3, 4.9, 
7.5, 
4.0, 8.2, 7.0, 
7.8, 
7.1, 
av. 7.2 
0.00388 sec. 
800 ft./sec. 
33, 
30, 35, 33, 33, 
33, 
33, 31, 32, 34. 
av. 32.7 
0.0176 sec. 
810 ft./sec. 
The agreement between the velocities as calculated from the average time 
interval with the keys 3 ft. apart and 14 ft. apart is better than one might 
expect from a small number of observations, considering the individual varia- 
tions. At 14 ft. between the keys the probable error of a single reading was 
only about 3 % . I believe that the moment of inertia of the keys can be further 
reduced and that thereby the errors would be reduced still more. 
The advantage that this apparatus has over clock and chronograph, aside 
from simplicity and ease of manipulation, is that there is no lag in the record 
arising from self induction or heavy moving parts. An electro-magnetic mark- 
ing device must show a retardation from these causes. 
A further adaptation that commends itself is in the measuring of small 
intervals of time in psychological research, such for example as the measure- 
ment of reactance time. The following is a set of observations on the reactance 
time of the motor auditory type. Each scale division represents 0.0039 seconds, 
deflections 
30 scale divisions 
27 
33 
27 
.28 
30 
33 
28.5 
31 
29.7 
av. 29.7 scale divisions or 0.1158 seconds 
The above shows a mean probable variation 0.007 second, which variation is 
probably entirely in the observer and not in the apparatus inherently. 
