the Pendulum vibrating Seconds in the Latitude of London. 175 
of one-tenth of an inch... The measurement of the same inch 
was repeated ten times at different periods, the microscopes be- 
ing previously adjusted anew each time for parallax. The mean 
results of such measurements are as follows : 
Divisions of the Micrometer to 1-10th of an Inch. 
2335-00 2335°75 
2333°75 2338°30 
2337°595 2335°85 
2337°32 2337°85 
2334°50 
233690 _. | Mean 2336°277 
Hence the value of one division of the micrometer appears to be 
gzrivz Of an inch. 
In the course of these measurements, differences occurred for 
which I was at a loss to account; but at length it appeared that 
they were to be attributed to remaining parallax ; for, whatever, 
care be taken in adjusting the microscope, it is scarcely’ possible 
to bring the image of the object precisely in the same plane with 
the threads, and the image will consequently be of various di- 
mensions, according to its distance from this plane. Unless, 
therefore, the most minute attention be paid to the adjustment 
for parallax, the error arising from this cause will be consider- 
able; and I may here remark, that I believe the difficulty of 
bringing the image into the plane of the threads, to be the source 
of by far the most serious errors to which measurements by 
means of microscopes are liable. 
I had now to examine the quality of the threads of the mi- 
crometer screw. For this purpose, two fine lines were drawn 
near each other on.a piece of brass; and the micrometer being. 
turned back as far as it would go, the distance of the lines was. 
carefully measured; and this was repeated, proceeding through 
the whole length of the screw, always advancing the micrometer 
one revolution previous to each successive measurement. The. 
result of this severe test will best appear by giving the numbers 
themselves. 
Divisions of the Micrometer. 
502-0 502°5 501:0 
5015 | 502 0 502°5 
501-0 502-0 501°0 
502-0 502°5 e 500-0 
5015 501°0 500-0 
502-0 5015 | 500°5 
502-0) 502°0 Mean 501°5 
The mean is 501°5, and the greatest difference from the mean 
ouly one division, amounting to yyzy5 of an inch, a degree of 
accuracy 
