234 
MINUTES OF PROCEEDINGS OF 
mitted compression seems to approximate much, more nearly to that due to 
the tension. 
Every ring, after it has been shrunk on, has its thickness altered (like a 
stretched piece of india-rubber), and the amount of alteration is not always 
uniform or certainly predictable. As is the case in all extended prisms of 
a metal, its density is diminished, and with that its compressibility orthogonal 
to the line of tension is increased. Hence the next outer ring, if previously 
bored to a size to give an assigned tension (on the supposition of unaltered 
thickness and physical condition), gives less than required. 
In large rings, errors in dimensions below those capable of being always 
avoided in the best and most careful practice, materially affect the result. 
In very large rings, differences in atmospheric temperature, in the workshop 
or outside it, may have sensible effects. Erom these and other such con¬ 
siderations, Sir William Armstrong, with his habitual acumen, soon saw 
that any attempt at great precision in apportioning the strains upon the 
successive rings was practically impossible, and publicly stated in 1860 
(Proc. Ins. C.E., Yol. XIX. p. 419) that in his guns “ the outer layers 
and rings were not put on with any calculated degree of tension; they were 
simply applied with a sufficient difference of diameter to secure effective 
shrinkage.” 
As respects the 36-inch mortars, I did not so completely cast aside theory. 
What I actually did do was as follows:—Having fixed, partly by theory 
partly by judgment, the assumed maximum strain of discharge and the 
possible maximum excess of temperature that might arise between the 
interior and the exterior of the mortar; and ascertained, by a few 
tolerably accurate experiments made at the works, Blackwall, most of the 
disturbing elements that hinder intended tensions and compressions in a 
number of superposed rings from being precisely attainable in practice; I 
then, starting from the middle point in the thickness of the chase, deter¬ 
mined the theoretic compressions within and tensions without that, for each 
of the plies of rings, and the successive differences of dimensions which, if 
given the rings, should upon that theoretic view produce these respectively; 
I then modified the results empirically, or by a mere exercise of judgment, 
in every case, so as to eliminate at .least some of the dirturbances, and 
more or less considerably to increase the tensions and compressions. 
In this I chiefly had regard to—1st, presumable errors in workmanship; 
2nd, stretching of the substance of the ring reducing its assumed grip; 
3rd, alterations in the cross sections of the rings after shrinking on. 
The dimensions when cold— i.e., at the atmospheric temperature of the 
time—inside and outside for every ply, were then tabulated, and steel 
gauges made by this, to which the rings were bored or turned. The whole 
of the rings were heated to a low red heat, visible in daylight, and were 
permitted to adjust themselves, being cooled rapidly by a stream of cold 
water, to prevent another source of disturbance—viz., the heating of the 
ply in the act of being hooped, by heat radiated first and then conducted 
to it from the red hot ply in the act of being superposed. 
With every precaution taken, some three or four of the rings, when struck 
by a light hammer, were found to be too loose, and had to be cut off and 
readjusted; and in one or two others it turned out that the want of grip 
was due to defective welds. The result of the operation has been that. 
