214 
PROCEEDINGS OF SECTION A. 
made stops, so that it cannot move from the position in which it is 
held. Means of adjustment of this plate-holder would be provided, 
and also the means for fixing it rigidly when it is adjusted. 
We now proceed to test the collimation of the camera, and wo 
assume that the photographic lens has been set as truly as possible at 
right angles to its tube, and therefore, since it turns on the same axes 
as the telescope which has been adjusted in collimation, it will in the 
two positions of the instrument record the star on exactly the same 
point of the plate, except there be any unknown errors of collimation 
or flexure due to the position of the telescope in declination, and 
differing from those found to exist in the horizontal and vertical 
positions in which it was tested. During the time the plate was 
exposed to the star the observer has constantly observed the star 
on the wire, and if the bisection has not been perfect he has instantly 
corrected it. So that the mean position is one of accurate bisection 
of the star, and experience shows that this is possible, because 
perfectly round photographic star images are easily obtained, and it is 
obvious that such a bisection of the star is immeasurably better than 
that of the flying shot method unavoidable in the ordinary transit 
circle. 
If the two images are not superposed then the centre of the spot 
made by the two images may be safely assumed to be the mean of the 
unknown errors, and if it does not wholly eliminate them it must at 
least be a more accurate position for the star than that obtained in an 
ordinary transit circle, which does not permit of the determination of 
these errors by any direct method, except in so far as the reversible 
transit circle of moderate dimensions provides for it. But the star 
camera form of reverser works much more smoothly than the other 
form, and has the great advantage of reversing not only the telescope 
but also the stand itself, and at the same time avoids jars to the 
instrument, and provides for a far more accurate bisection of the star 
as we have seen. 
In taking observations or rather photographs by this instrument 
two observers are necessary, and the work would be better and much 
more rapidly done by three — one to observe the star bisected by tlie 
wires, another to read the microscopes on each declination circle in 
each position of the polar axis, and the third to read and record on a 
chronograph the time of transit. It has already been mentioned that 
the graduations for 15 minutes of time on each side of the zero, i. e . , 
the point indicating that the telescope was in the plane of the meri- 
dian, and the determination of that point by observation of a mercury 
reflector, have also been stated. It remains now to mention that 
BA graduations or marks should be at 10 seconds intervals ; this is 
quite possible since the BA wheel is 3 feet in diameter. Hence seven 
of these marks would pass under the microscope each minute ; the 
microscope is powerful enough to enable tlie observer to see easily 
of a second of time, and the parallel wires in the microscope 
should be exactly suited to the division lines. A point on the gradua- 
tions would be chosen so that the observation begun there would be 
soon enough to admit of reversal and picking up the corresponding 
divisions on the other side of the meridian. Since there would only 
be a few of these division lines in use, it would be possible to deter- 
mine with great accuracy their relation to the meridian line. 
