OBSERVATION OF SUDDEN PHENOMENA. 
47 
the first image on the retina must be that of the star as 
seen in some particular one of the four quadrants. 
Going a little farther, we will now suppose each of the four 
quadrants, which in turn correspond to quarter seconds, to 
be divided into five parts, so that the whole circle is divided 
into twenty. All the observer has to say is in which quad¬ 
rant and in which subdivision of the quadrant the star ap¬ 
pears, to say in which twentieth of the second (or other 
brief cycle) it emerged. 
The reticule I have just described is fixed in the focus of 
the eye-piece and does not revolve. What does revolve is a 
minute double prism of total reflection just before the reti¬ 
cule, the middle of whose reflecting face lies in the optical 
axis, and by whose means the optical axis is twice broken 
at a right angle, so that when the telescope is directed at a 
star the image of the star is not seen at the center of the field, 
but on one side of it. If the prism is revolved, the star must 
appear to revolve in a circle whose radius is nearly that of 
the side of the prism. 
The little prism is turned by a small piece of watch-work, 
but it is not at all necessary that this should be exact, since 
all we demand is that the rate shall be constant during a 
second or so-—a condition easily secured with the most ordi¬ 
nary mechanism. 
The sketch and the apparatus exhibited sufficiently indi¬ 
cate, I think, the simple means by which this is brought 
about. 
Figure 1 is a section one-half full size. A A A is the outer 
tube, which can be fitted, if desired, into the eye end of a tele¬ 
scope. b b is the inner tube, resting on friction-wheels //, 
revolved by the clock-work c about once a second, and 
recording the time at which a key in the observer’s hand 
may be pressed to indicate the particular second. This rec¬ 
ord may be made electrically by the wires w w on a chrono¬ 
graph, or more simply and directly on a little attached dial 
like that of a recording stop-watch. 
p is the prism of double total reflection, r r is the posi- 
