90 
inch at 415°. This may, however, be accounted for from 
the increased manipulation of rolling where the fibre is 
drawn and elongated to a much greater extent than in 
plates. This does not, however, affect, to any great extent, 
the ratio of compression and extension as regards the effects 
of temperature, although I should he inclined to take the 
experiments on plates before that on bars, as analogous to 
the broken tire, which, it must he borne in mind, is without 
weld and perfectly homogeneous. 
The danger arising from broken tires does not, according 
to my opinion, arise so much from changes of temperature 
as from the practice of heating them to a dull red heat, and 
shrinking them on to the rim of the wheels. This, I believe, 
is the general practice, and the unequal, and in some cases, 
the severe, strains to which they are subject has a direct 
tendency to break the tires. 
To show how easily this may be effected, let us suppose 
that a tire, two feet six inches or three feet diameter, is 
shrunk on to a wheel one-tenth of an inch larger than the 
tire, it then follows that the tire in cooling must be elonga- 
ted to that extent, with a strain, equivalent to the force 
of the shrinkage, and calculated to produce that amount of 
molecular disturbance. It may be more or it may be less, 
but supposing the strain to be one-half or three-fourths of 
that which would break the tire, it then follows that the 
constant action of its irregular motion on the rails must 
ultimately lead to fracture* 
I am not surprised that this should be the case, as most, 
if not the whole, of railway tires, excepting those on engines 
and tenders, are not turned but selected by hand, heated 
and shrunk upon the wheels with every degree of tension, 
as suits the 'convenience of the workman. So long as this 
process is pursued, the public will be exposed to the risk of 
broken tires. 
# Prom long continued action under strain, it lias been proved that it is 
only a question of time when rupture takes place as repeated increased and 
diminished changes with the same load ultimately leads to fracture. 
