(624.) The general conclusion from these experiments is, that for permanent structures the maximum working load on rolled bars of wrought iron should not exceed 8 tons per square inch, which is nearly one-third of the mean ultimate cohesive strength of British iron, namely, 25.7 tons per square inch, as shown by Table 1. Within the limit of 8 or 9 tons, the extensions are practically simply proportional to the strains, and may be taken approximately at 00008 of the length per ton. With heavier strains, the extensions increase more rapidly than the load, and with all such strains are still further extended by time to an extent varying considerably with the ductility of the particular specimen of iron.

(625.) Another important practical lesson may be drawn from this investigation, namely, that where the length is fixed, whatever initial strain may be put on wrought iron, the permanent strain after a certain time will not exceed 8 or 9 tons per square inch. For instance, say that a wrought-iron hoop or wheel tire is shrunk hot on a cast-iron wheel, &c., as in Fig. 25 and (90), so that when cold it shrinks to such an extent as to yield at first say 20 tons per square inch; but that strain would not be permanent; the bar would immediately begin to stretch, and by so doing relieve the pressure upon it, at first very rapidly, afterwards, as the strain was reduced, more and more slowly, until after a long period it became 8 or 9 tons only, although the extent to which it was originally stretched was that due to 20 tons, and that strain was really obtained for a few seconds (90).

(626.) "Compression of Wrought Iron."-Experiments were made by Mr. Hodgkinson on two bars of wrought iron, each about 1 inch square and 10 feet long:-they were enclosed in a frame to prevent lateral flexure, in the same way as with the cast-iron bars in (607). The results are given in Table 97, but the experimental strains on the bars (which were rather more than one inch square) are reduced to equivalent strains per square inch, and the observed compressions are reduced to parts of the length of the bar.

The compressions for even tons were obtained by interpolating between the greater and lesser experimental numbers.

The mean compression of the two bars is given by Table 98, and in cols. 4, 5, reduced results are given, which may be taken as sufficiently correct for ordinary purposes.

=

(627.) "Effect of Time."-The effect of time in increasing the amount of compression was observed with one of the bars only, and that not before a strain of 30,858 lbs., or 13.77 tons, was applied: the influence of time was then found to be very great, the compression being increased 002175·00158333 1.37, or 37 per cent. in an hour; 48 per cent. in hour; and 73 per cent. in 1 hour. In all probability, time would have been found influential with compressive strains of 8 or 10 tons per inch if observations had been made, and we may infer that for permanent structures the maximum working compressive strain should not exceed 8 tons per square inch, being the same as we found for the Tensile strain (625).

(628.) "Comparative Extension and Compression of Wrought Iron."-We found (616) that with small strains, when defect of elasticity was uninfluential, cast iron yielded more to compressive than to equivalent tensile strains. A similar comparison of the elasticity of wrought iron leads yet more clearly to the same result, as shown by Table 91; the elasticity of wrought iron is so nearly perfect that the ratio is nearly the same with all strains up to 10 or 11 tons, the mean ratio in col. 7 from 1 to 11 tons is 1.236.

The length of rods and pillars of wrought iron that would be shortened 1 inch by different tensile and compressive strains is given by Table 93, and will suffice to give a general idea where it is desired to avoid the trouble of exact calculation.

(629.) "Comparative Extension of Cast and Wrought Iron."These two important materials are frequently combined in structures, and the differences in their elasticity cause unequal strains under circumstances where perfect equality might have been expected.

We will first consider the relative extensions of cast and wrought iron under the same tensile strain :-here the maximum strain must of course be limited by the strength of the weaker material of the two, namely, cast iron, which may be taken at

TABLE 97.-Of the COMPRESSION of WROUGHT IRON by CRUSHING STRAINS: from Direct Experiments on two Bars 1 Inch Square.

TABLE 97. Of the COMPRESSION of WROUGHT IRON by CRUSHING STRAINS continued.

7 tons per square inch. In Table 99 the cols. 2 and 3 have been taken from cols. 4, 4, of Tables 88, 96, and col. 4 gives the ratio which increases regularly throughout, this fact being due to defect of elasticity in the cast iron as opposed to the almost perfect elasticity of wrought iron, the ratio rising from 2.066 with 1 ton to 3.008 with 7 tons per square inch. Thus, even with so low a strain as 1 ton, the extension of cast iron is more than double that of wrought iron.

(630.) But this statement does not give a clear idea of the effect of the unequal resistance to the same extension which happens in those numerous cases where the two materials are so combined that a given load must of necessity stretch them both to the same extent. Thus, in Fig. 123, let A be a rod of cast iron one square inch in area, and B a similar one of wrought iron stretched simultaneously and of necessity to the same extent by the weight W. Now, with say 4 tons per square inch on A the extension by col. 4 of Table 88 would be ⚫00077282,