696 EAEL OF BOSSE ON THE CONSTEUCTION OF SPECULA OF 6-FEET APEETUEE, 
proper height at each side of the board. The surface of the pitch is wiped dry, and a 
thin stratum of the hard resinous composition in powder is sifted on. A large flat-iron, 
red-hot, is then passed over at a distance of 3 or 4 inches, and so slowly as just to fuse 
the resinous powder without making any change in its composition. The pitch, so pre- 
pared, is cut into squares of the proper size, and thrown into cold water till required. 
The polisher, warmed to about 80°, is then brushed over with very soft pitch, and, when 
the temperature has fallen to about 65°, the square pieces of pitch are arranged in their 
places and soon become quite fast. The whole of this operation requires little expe- 
rience, and can be managed by common workmen — a great advantage. It has, however, 
this disadvantage, that the pitch is somewhat thicker than we should wish. In the 
6-feet polisher the squares are 2^ inches ; and although the soft pitch in the circular 
grooves will no doubt yield a little, still we have a larger continuous surface than we 
had by the original process, and therefore the pitch requires to be thicker. The reason 
why with the long transverse strokes the pitch must be so thin is evidently this, that 
the polisher passes so far beyond the edge of the speculum. If we coat a polisher with 
pitch alone and of some thickness in the ordinary way, and then proceed to polish, we 
shall find that, if at any time we stop the machinery for a few moments when the 
polisher is at the extremity of the stroke, the pitch will change its figure. The change 
of course will be less as the stroke is shorter ; but by prolonging the time, even with a very 
short stroke it will still be perceptible. So considerable is the change of figure under 
such circumstances, that after some time a distinct mark will be made by the edge of the 
speculum, and the projecting portions of the pitch becoming comparatively protuberant, 
unusual force will be required to efiect the next stroke. These continual changes of 
figure, slight as they may be, will produce excessive action on the outer portions of the 
speculum. To meet this evil, if we diminish the length of the strokes much, we impair 
the self-correcting action to which we are mainly indebted for success. To explain this, 
let us suppose the throw of the first eccentric, B (see figure in ‘ Transactions’ for 1840), 
to be reduced to a small quantity, and the action of the second eccentric, G, to be reduced 
in the same proportion, the speculum continuing to rotate ; if the polisher and speculum 
are not truly portions of the same sphere, there will be unequal action at the centre or 
edge of the speculum, according as the polisher is more or less convex than it ought to 
be. In the first case, a depression will be formed at the centre of the speculum of a 
diameter proportional to the throw of the eccentrics ; in the second case there will be 
an annular depression at the edge of the speculum. It is plain that the speculum 
cannot be restored to truth till the remainder of the surface has been lowered to the 
depth of the depression : this, however, will not be accomplished in practice if unequal 
action is continued even for a very short time. It may be thought that rigid identity 
of figure might be secured in the first instance ; but this is practically impossible : the 
rouge cannot be distributed with perfect regularity ; besides, as the temperature of the 
polisher varies, so does its radius of curvature. But even if perfect coincidence was 
secured at the beginning, it would not long continue. With very small motions the 
