. 
| 
NATURE 
427 
‘s apply to the riveting of iron apply equally to steel; that is to 
t 
> 
, 
say, that the total shearing area of the rivets must be the same, 
or rather must not be less, than the sectional area of the bar 
riveted... 
We know from established mechanical laws that the limiting 
spans of structures vary directly as the strength of the material 
employed in their construction when the proportion of depth to 
a and all other circumstances remain the same. We know 
so that, taking an ordinary form of open wrought-iron detached 
_ girder (as, for example, when the depth is one-fourteenth of the 
_ span), the limiting span in iron, with a strain of 5 tons to the 
inch upon the metal, is about 600 ft. ; and it follows that a steel 
girder of like proportions, capable of bearing S tons to the inch, 
would have theoretically a limiting span of 960 ft. 
This theoretical limiting span of 960 ft. would, however, be 
reduced by some practical considerations connected with the 
minimum thickness of metal employed in certain parts, and it 
would, in effect, become about goo ft. forja girder of the before- 
__ mentioned construction and proportions. 
The knowledge of the limiting span of a structure, as has 
been explained elsewhere, enables us to estimate very quickly, 
and with close approximation to the truth, the weight of girders 
required to carry given loads over given spans ; and although 
the limiting spans vary with every form of structure, we can 
obtain an idea of the effect of introducing steel by the relative 
weights of steel and iron required in girders of the kind above 
mentioned, 
Assuming a load in addition to the weight of the girder of one 
ton to the foot, the relative weights under these conditions 
would be as follows :— 
Weight of steel Weight of iron 
Span. girder. girder. 
tons. tons. 
200 57 100 
300 150 300 
20 800 
400 aod 
It is not alone in the relative 
the advantage of the stronger material is important, but with 
steel we shall be enabled to cross openings which are absolutely 
impracticable in iron. 
t will naturally be asked why it is that steel is not used in 
these structures, if such manifest advantages would result from 
its employment. 
The reason is twofold :— 
Ist. There is a want Of confidence as to the reliability of steel 
in regard to its toughness and its power to resist fracture from 
sudden strain. 
2nd. Steel is produced of various qualities, and we do not 
possess the means, without elaborate testing, of knowing whether 
the article presented to us is of the required quality for structural 
purposes. A third reason, arising probably out of those before 
mentioned, is found in the fact that in the regulations of the 
Board of Trade relative to railway structures, although rules are 
given for the employment of cast-iron and wrought-iron, steel 
has not, up to the present time, been recognised or provided for. 
Now, as regards the question of toughness and malleability, 
and referring again to Mr. Kirkaldy’s experiments, it appears 
that in the tests of “‘ Bessemer steel” 18 samples were tried under 
tensile strain, the length of the samples being in round numbers 
50 in. and the diameter 1°382 in, ; and that when these were 
subjected to ultimate strain, the elongation at the moment of 
fracture was in the most brittle example 2$ in., but generally 
varied from 44 to 9} inches. a 
In the experiments on transverse strain, in which the bars 
were nearly 2 in. square and only 20 in. between the points of 
_support, all the ‘‘ Bessemer steel” samples, except two, bent 
6 in, without any crack. Again, in the experiments made by 
the Committee on bars 14 ft. long and 1} in. in diameter, out of 
20 bars of the milder quality of steel, 16 extended more than 
8 in., and of these ro extended more than 12in. .. . 
* The treatment by comparison is especially important where 
metal is required in large masses and/of great ductility because 
the larger the mas. and the greater the ductility, the larger and 
more numerous are the air-cells, and the effect of the pressure 
is to completely close these cells and render the metal perfectly 
solid. 
By this process mild steel can be made with a strength of 40 
tons to the inch, having a degree of ductility equal to that of 
the best iron, 
The more highly carbonised qualities show a decrease of 
ductility somewhat in the same ratio as the strength increases. 
weight or in the relative cost that 
Without going into the numerous achievements of Sir Joseph 
Whitworth, resulting from the employment of steel, in connection 
with the extreme accuracy of workmanship produced at his 
works, or doing more than mention the flat-ended steel shot and 
shell which pass through iron plates when fired obliquely or 
penetrate ships’ sides below the level of the water, I would call 
attention to those applications of steel which bear upon its 
strength and toughness. 
In the first place, there are small arms made entirely of steel, 
of wonderful range and accuracy, capable of penetrating 34 half 
inch planks, which is about three times the penetrating power 
of the Enfield rifle. 
Secondly, there are the large guns, also entirely of steel, 
throwing projectiles from 250 Ibs. to 310 Ibs. in weight, and 
burning from 40 to 50 lbs. of powder at a charge, with which a 
range of nearly 64 miles is obtained. 
In both these cases the degree of strength and toughness re- 
quired in the metal is much greater than is necessary for engi- 
neering structures. 
It is unnecessary to occupy more time in multiplying examples 
of the toughness of steel. It is well known to manufacturers, 
and must also be well known to many others here present, that 
steel of the strength of 33 or 36 tons per inch can be made, 
and is made in large quantities at moderate price, possessing 
all the toughness and malleability required in engineering struc- 
tures. 
I will proceed, therefore, to the second part of the subject 
—namely, the want of means of knowing that a given sample 
of steel is of the quality suited for structural purposes. 
With most other metals chemical analysis is in itself a com- 
piete and sufficient test of quality, but in steel it is not so, The 
toughness of steel may be altered by sudden cooling; and 
although the effect of this operation, and generally the effects of 
tempering, are greater when the quantity of carbon is consider- 
able, yet it acts more or less in the mild qualities of steel; so 
that we cannot rely entirely on the aid of the chemist, but must 
fall back on mechanical tests. And in point of fact, seeing that 
the qualities required are mechanical, it is no more than reason- 
able that the test should be mechanical ; for this includes not 
only the test of material but of workmanship. 
Now there are two descriptions of mechanical testing, which 
may be distinguished as destructive and non-destructive—the 
one being beyond and the other within the elastic limit of the 
material. The destructive test is that usually applied to a part 
of an article manufactured, as, for example, a piece cut offa 
boiler plate and tested by absolute rupture, or by bending or 
otherwise, whereby the strength and quality of the material in 
the plate is known, 
The non-destructive test is that usually applied to the finished 
work, as in the test of a boiler by hydraulic pressure, or the test- 
ing of a gun by the proof-charge. The strain in this case is 
made greater than that which will arise in the daily use of the 
article, but is not so greatly in excess as to be beyond the elastic 
limit of the material. 
As regards engineering structures, this second test is easy of 
aplication ; but it affords no sufficient criterion that the metal 
possesses that degree of toughness necessary to resist the action 
of sudden strains. 
It may be said that engineers may ascertain for themselves, by 
inspection and testing at the works, that they are being supplied 
with the material that they require ; but assuming that the tests 
and mode of testing were in all respects satisfactory to them, and 
that the metal supplied was of the right quality, we have still to 
comply with the conditions of the Act for the Regulation of 
Railways, and we must satisfy the Government Inspector. 
It isnot to be supposed that he can attend all the required 
tests at the works ; and the question remains, how is the In- 
specting Officer of the Board of Trade to be enabled to distin- 
guish the quality of metal in a finished bridge, when he is called 
upon to give a certificate that it is safe for public traffic? 
If we could adduce clear and distinct evidence that the metal 
used for a bridge was of a quality which would bear 8 tons to 
the inch with as much safety as common iron can bear 5 tons, 
there can be no reasonable doubt that the Board of Trade would 
make suitable provision in its regulations for the employment of 
such material. 
The difficulty lies in the want of something whereby the quality 
of the metal may be known and relied upon with confidence by 
others besides those who made the article. 
In gold and silver this is accomplished by the stamp put upon 
