17S An Account of Experiments for determining the Length of 
the first line from which the measurements were to be taken. 
These pieces were marked A, a, and B, 0. 
The pieces A and a, being placed with their edges in contact, 
in which position they were kept by the pressure of a spring, the 
distance between the fine lines first drawn was carefully measured 
with the micrometer, and from a mean of eight observations, the 
greatest difference between which did not exceed one division, 
was found to be 329-09 divisions. 
~ The same was done with the pieces B and b, and the distance 
of the lines from a mean of sixteen observations appeared to 
be 366°96 divisions. 
The knife edges being adjusted as nearly as possible parallel 
to each other, the pieces A, a, and B, b, were placed in contact 
with those parts of the knife edges on either side of the bar, on 
which the vibrations were to be performed, and were retained in 
their places by the pressure of slight springs attached to the 
mahogany case. 
The microscopes were now breught over the pieces A and a, so 
as for the lines before described to bisect the cross threads, when 
the division of the micrometer was noted. 
The same was done with the pieces B and b; and the division. 
of the micrometer was also registered. 
The pendulum being removed, the standard scale* was placed 
beneath the micrometer ; ; and its zero being made to bisect the 
angles of the fixed microscope, the cross threads of the micro- 
meter microscope were brought to 39:4 of the scale, and the 
revolutions and parts of the micrometer were noted, 
From these data, and the respective distances of the lines on 
A and a, and on B and b, when the pieces were in contact, the 
distance of the knife edges on either side of the 'bar may be 
readily obtained, and dhe mean being taken, will obviously cor- 
rect any error arising from a want of perfeet parallelism in the 
knife edges. 
It is very generally believed that measurements from a knife 
edge, or from a line terminating a surface, are liable to much 
uncertainty from what has been called irradiation, or indistinct- 
ness of the image. But this is by no means the fact ; for, if the 
reflection of light from the knife edge be prevented, and it be 
viewed on a white ground, it may be made to bisect the cross 
threads of the microscope, with nearly the same precision as 
could be attained by the use ofaline. There is, however, a cor- 
rection necessary to be applied in this case, and I shall proceed 
to describe the method employed for ascertaining its amount. 
A slip of writing-paper was pasted on the mahogany case, under 
each knife edge, extending beyond it about the tenth of an inch, 
* The scate constantly referred to is Sir George Shuckburgh’s ae 
an 
