376 
ME. WAEEEN DE LA EUE ON THE 
tions were found not to alter materially for several months, during which they were from 
time to time tested. 
The next thing to be ascertained was the rectangularity of the two slides, which was 
done in this way. By means of the upper circle D, one of the ruled lines on the glass 
plate was made to coincide with a wire of the microscope after the lower ch’cle C had 
been clamped to read 360°. The slide B remaining central, slide A was undamped and 
drawn out to the full extent ; any deviation between the line on the glass plate and 
the centre of the cross in the microscope was then corrected by moving the tangent 
screw (?', and, the slide being again pushed back to the centre, a few trials soon brought 
about an exact coincidence of the line on the plate and the wme of the microscope during 
the travelling of the slide A. The vernier V was next read off accimately, and the slide 
A brought to its normal position, 2 •0025 inches. The circle C was then moved tlii-ough 
exactly 90°, and clamped securely, and the slide B, having been undamped, was pulled 
out to the extent of its path. If the same line on the glass plate maintained its coin- 
cidence with the cross of the microscope during the travelling of slide B, the slides were 
necessarily at right angles ; if not, the error was corrected by moving the slide B, with 
respect to A, by means of the adjusting-screws a'"'. If the deviation was considerable, 
it became necessary, after its rectification, to re-ascertain the normal central position of 
slides A and B. The greatest deviation from rectangularity amounted, after final adjust- 
ment, to 0° 0' 10", a quantity which could in no way affect the measurements. 
The thickness of the wire of the microscope was ascertained to be 0*000 3 inch, amount- 
ing to about 0"-15. The measurements could not generally be made nearer than 0-001 
inch, equal to 0"-5, and in most cases the centre of the wire was brought as nearly as 
possible to coincidence with the point to be measured ; this minute correction conse- 
quently has not been applied. 
The measurement of the photographs was effected in the following manner : — In the 
first place, the slide B was set at its central point 1'981 inch, and the slide A pulled out 
so far as was judged necessary to bring the periphery of the sun (or moon, as the 
case might be) under the centre of the cross. The photograph to be measured was 
then laid on the supports, due regard being paid to the position of the wires depicted 
thereon, so that they might coincide with the corresponding divisions on the circle C, 
that is, I. with 45°, II. with 135°, III. with 225°, and IV. with 315°. The photograph 
was next centered by means of the four screws d', and the requisite movement of the 
slide A by means of its screw a", the slide B remaining at rest. When the periphery 
of the sun, notwithstanding its somewhat irregular outline, maintained, during the rota- 
tion of the circle C, a general coincidence with the centre of the microscope, the vernier 
a' was read off ; this reading, minus the reading of the central position of A, namely, 
2'0025, gave the measurement of the radius of the sun. The slide A was then un- 
damped and traversed right across, so as to bring the limb of the sun, after passmg to 
the other side of the centre of the instrument, into coincidence with the cross when the 
circle C was rotated, and the vernier a! again read off. The difference between the fii'st 
and second readings gave the diameter of the sun-picture. Thus, let 
