roughing is not always carried out exactly as
described, and will be referred to again.

(4) The glass being approximately of the proper radius of curvature on
one side, it is reversed on the chuck and the same process gone
through on the other side. After this the glass is usually dismounted
from the lathe and mounted by cement on a pedestal, which is merely a
wooden stand with a heavy foot, so that the glass may be held
conveniently for the workman. Sometimes a pedestal about four feet
high is fixed in the floor of the room, so that the workman engaged in
grinding the lens may walk round and round it to secure uniformity.
For ordinary purposes, however, a short pedestal may be placed on a
table and rotated from time to time by hand, the operator sitting down
to his work.

Rough iron or brass tools do not succeed for fine grinding--i.e.
grinding with fine emery, because particles of emery become embedded
in the metal so tightly that they cannot be got out by any ordinary
cleaning. If we have been using emery passing say a sieve with 60
threads to the inch, and then go on to some passing say 100 threads to
the inch, a few of the coarser particles will adhere to the "tool",
and go on cutting and scratching all the time grinding by means of the
finer emery is in progress.

To get over this it is usual to use a rather different kind of
grinding tool. A very good kind is made by cementing small squares of
glass (say up to half an inch on the side), on to a disc of slate
slightly smaller than the lens surface to be formed (Fig. 51). The
glass-slate tool is then "roughed" just like the lens surface, but, of
course, if the lens has been roughed "convex" the tool must be roughed
"concave".

The "roughed" tool is then used to gradually improve the fineness of
grinding of the glass. For this purpose grinding by hand is resorted
to, the tool and lens being supplied continually with finer and finer
emery. Fig. 52 gives an idea of the way in which the tool is moved
across the glass surface. Very little pressure is required. The tool
is carried in small circular sweeps round and round the lens, so that
the centre of the tool describes a many-looped curve on the lens
surface. The tool must be allowed to rotate about its own axis; and
the lens and pedestal must also be rotated from time to time.

Every few minutes the circular strokes are interrupted, and simple,
straight, transverse strokes taken. In no case (exce