360 Mr A. J. Avie on the Expansion of Stone 
ment and back again to the chimney passed through holes in 
the screen-boards. A detached thermometer was hung over the 
instrument, and it was always kept at the same height by admit- 
ting more or less air by the window. In order to have the same 
direction of light during each experiment, I employed two small 
lamps, placed about a foot from the instrument, and their light 
was thrown in at the windows by lenses and reflectors. The re- 
flectors were fixed on soft wires, so that they could be bent to re- 
flect a bright light from the bottom of the lines on the studs. All 
the parts of the instrument were made as heavy as consistent with 
convenience, to prevent their being affected by sudden changes 
of temperature. 
The rods of which the expansions were determined varied from 
half an inch to an inch square, and the length of twenty-three 
inches was carefully laid off on silver-studs, which were fixed into 
holes drilled inthe rods, the centres of the holes being twenty-three 
inches apart. It was necessary to divide the head of the upper 
stud into fiftieths of an inch, in order to determine the value of 
the micrometer for each experiment, as it was liable to a little 
variation, from the impossibility of adjusting it to exactly the 
same focus, which it had when it was adjusted to a fixed value. 
To guard against this error arising from alteration of the focal 
distance, I read the micrometer for 2; of an inch on the stud in 
the rod which was to be heated ; and this was repeated at each 
experiment made on the rod, after every thing had been adjust- 
ed. In placing the rod in the instrument a socket was tied to 
the lower end below the lower stud, and the socket was fitted so 
as to slip tightly upon the top of a square head, on the end of a 
pin P, Fig. 3, moyed by the lever and screw G, Fig. 1. This served 
to raise and depress the rod, and was very convenient in fixing it 
in the proper place, for the under microscope. In Fig. 3, A is the 
lower end of the rod, S the under stud in it, B the socket, which 
is tied to A, by the piece on the front, so that it was of no con- 
sequence what the thickness of the rod might be, as the stud was 
