110 TRANSACTIONS OF THE [JAN. 16, 
exposure is always obtained, and is not much shorter than the 
primary exposure. My experience leads me to believe that a 
large proportion of the instantaneous shutters in the market are 
worthless, owing to this defect. The remedy for it is a large 
overlap to cover, ora catch to prevent the rebound. Neither 
of these precautions is usually provided. The effect of this 
double exposure upon motionless objects is to give a blurred pic- 
ture, usually attributed to inaccurate focussing. A large pro- 
portion of failures, and the loss of many plates probably occur 
from this cause. 
Fig. 13 was obtained with a beam of normal, and Fig. 14 with 
a beam of condensed sunlight, and both with the quickest action 
of the shutter that had been employed with that lens. In both 
the first and second pair of these four records, the defect of Londe’s 
method, namely, its dependence on the intensity of the light, is 
at once apparent. ‘To obviate this variable element, I removed 
the lens, and by so locating the camera that the beam from the 
meniscus, Fig. 1, focussed on the back shutter, passing centrally 
through the shutter opening, which enlarged in the plane of vi- 
bration of the fork, I obtained, with the same exposure, repre- 
sented in Figs. 13 and 14, and a beam of normal sunlight, the 
record shown in Fig. 15. ‘This appears to give the true time of 
the exposure, and seems to reveal an accurate and simple method, 
especially adapted for determining the time of exposure of any 
diaphragm shutter whatever. ‘The velocity and expansion of 
the receiving film are matters of indifference. It is only neces- 
sary to divide the rate of the fork by the undulations obtained 
to arrive at the time of exposure which, in this case, is about 
<i}, of a second, obtained by dividing 331.25 by 24, a fair es- 
timate, perhaps, of the number of undulations found in Fig. 15, 
which suggests that for closer reading of the more rapid expo- 
sures, a fork of much higher rate may be advantageously em- 
ployed. 
Instead of a sliding shutter, a rapidly rotating disk may be 
employed to carry the sensitive film, which, as the weight is then 
immaterial, may be a plate and the correction for expansion be 
thus avoided. By means of a simple multiplying wheel I ob- 
tained the record given in Fig. 16, of a Forrest ‘‘due ratio” 
shutter at a slow exposure, which was chosen in order to obtain 
nearly a circular tracing. It is evident, however, that by re- 
volving the plate with sufficient rapidity the same result may be 
secured with any exposure. Rapid rotation may evidently be 
obtained by electro-magnetic devices, and either the duration of 
the exposure, or with a little calculation, the rate of rotation of 
the plate may be determined. Ifthe weight of a plate would 
