114 MODES OF MEASURING MINUTE INTERVALS OF TIME. 
produced when the eye is fixed on a swiftly revolving fly wheel, provided the axis of vi- 
sion follow for an instant the ascent or descent of the rim. It will thus happen that the 
ends of the arms are seen momentarily with considerable distinctness, so as to give to the 
revolution an appearance of irregularity, which does not belong toit. In observing steadily 
the current of a river, there is a natural tendency in the axis of vision to follow bubbles, 
patches of foam, or other objects floating with the stream, against which it requires an 
effort constantly repeated, to replace the axis in the same direction. This adds to the 
variety and life-like motion of the stream. If attention be given to a cascade, with the 
axis of vision directed towards its summit, there is the same natural tendency to let it de- 
scend with the masses of falling water, which are thus revealed individually in their pro- 
per form and dimensions, all disappearing as the eye glances upward again. Hence there 
is an apparently capricious variation of appearance in these and some analogous instances, 
which never can be caught in painting, for all representations of them must correspond to 
their aspect when the eye-axis is absolutely fixed, or when the moving masses are not in 
its focus of vision. The effect is obviously traceable to the longer duration given to one 
impression in the eye, while its axis follows the movement, or to its relief from a succeed- 
ing impression coming on to interfere with one already received. 
Another method of producing the same effect is by winking, or by rapid closing of the 
eyelid. The circumstances already mentioned as affecting the appearance of a waterfall, 
or other quick movements of masses in motion, are sometimes due to this influence. By 
thus preventing a subsequent impression from over-lapping and confusing one already re- 
ceived, an image before the mind is for a time stored up there, to be contemplated by it. 
If in marking the instant of bisection, by means of a signal given to a time-movement, it 
could be so arranged that the image in the state of bisection were instantly covered, then 
the examination of this last impression received from it, would decide as to the complete- 
ness of the bisection when the signal was given. This closing off of the image might be 
produced by the same finger-movement which sent the signal on to the time piece; or the 
eyelid and the finger might be trained to act in conjunction. Trial will readily show that 
different portions of the muscular frame act with sympathetic readiness under the same~ 
movement of the will. 
As to the time-movement to which, in such a mode of proceeding, a signal is to be given, 
it seems to have been assumed that there is a necessity for a perfectly uniform motion, or 
one such that signals given after equal times were found to be separated by equal spaces on 
the dial-face, or on whatever surface may receive the mark of the signal. This uniformity 
is evidently by no means indispensable. It is enough that the movement take place under 
some known law, so that from the measurement or register of the spaces passed over by the 
