proper place. (See Fig. 25.) There were, of
course, many of the wheels with small dish and long axles still in
use. Their treatment, however, when the day of change came, did not
vary from that of the short axle.

[Illustration: FIG. 24 and FIG. 25]

It had been the rule for some years that all axles should be turned
back 11/2 in. further than needed; but unfortunately the rule had not
been closely followed, and many were found not to be so turned. To
make the matter worse, quite a number of the wheels were found to have
been counterbored about 1/2 in. deep at the back end, and the axle
turned up to fit this counterbore; a good idea to prevent the running
in, in case the wheel worked loose, but bad from the standpoint of a
change of gauge. In such cases the wheels had to be started off before
the axle could be turned back, so that the wheels could be pushed on
in their proper position. (Fig. 26.)

[Illustration: FIG. 26]

If the work was done where they had a lathe large enough to swing a
pair of wheels, they were pressed off but half an inch, the wheels
swung in the lathe, the axles turned back 11/2 in., and the wheels then
pressed on 2 in. or 11/2 in. inside of their first position.

Where no large lathe was in use, the wheels came entirely off before
the axles could be turned back. The work in the former case was both
the quicker and the cheaper. Where the large lathes were used they
were either set down into the floor, so a pair of wheels would easily
roll into place, or a raised platform was put before the lathe, with
an incline up which the wheels were rolled and then taken to the
lathe. These arrangements were found much quicker and cheaper than to
hoist the wheels up, as is usually done.

In pressing the wheels on, where the axles had previously been turned
back, much trouble was at first experienced because of the rust that
had gathered upon the turned part behind the wheel, forming a ridge
over or upon which the wheel must be pushed. Some of the roads, at the
start, burst 10 or 15 per cent. of the wheels so pressed on. By
saturating this surface with coal oil, however, it was found that the
rust was easily removed and little trouble was had. It was found,
sometimes, that upon axles newly turned back a careless workman would
leave a ridge at the starting point of the turning. Frequently also
the axles were a little sprung, so that the new turning would be a
little scant upon one side when compared with t