As far as the mechanism of this strain is concerned it is the most simple of any of those above enumerated, but it is by no means the easiest to submit to experi ment, particularly in timber, because, if the force is not directly opposed to the fixed point, the fibres are liable to be destroyed by a twist or strain different from that we arc endeavouring to estimate; it is pro bably to this circumstance we must attribute the dis cordance observed in the results obtained by different experimenters; and it is for this reason we have thought it right to inform the reader of the means employed to assure accuracy in this respect in making the following experiments. Referring to Fig. 1. Plate AB represents one of the pieces whose strength is to he determined, its whole length being twelve inches; the length of each square three and a half inches, and the side of the square one and a half inches: the intermediate part, five inches, was turned in an excel lent lathe and by a good workman, and brought down in the centre to one-third or one fourth of an inch in diameter, its exact diameter being ascertained by winding a piece of silk ten times round the circum ference, which length divided by ten, gave the cir cumference, and hence the diameter was computed. The other cylindrical parts were each three-fourths of an inch in diameter. CC, DD, Fig. 2, represent two strong iron bars brought to the form shown in the drawing, GG, are two screws which are passed through the holes NH, in the bar DD, and are there screwed fast by the nuts II; EE are two semicircular collars rivetted, one to each bar, which, when the two are screwed together, form a circular plate, as repre sented in Fig. 4. The circular hollow parts ee, are three-fourths of an inch in diameter, so as to fit exact ly the larger part of the cylinder shown in Fig. 1. These bars, after being screwed together, were rested on the supports as in Fig. 4, and then brought out of winding and accurately adjusted to a horizontal posi tion by a spirit level.
The two iron boxes, MNO, M'N'O', Fig. 3, were made exactly to fit the square head B of Fig. 1, hav ing also two semi-circular holes at the top correctly fitted to the larger part of the cylinder: these were shut by passing the bolts through the holes NM, and were thus secured by the shears shown in Fig. 4. In making the experiments, the head A of Fig. 1, was placed above the collars LE, Fig. 2, the upper larger cylindrical part of Fig. 1 being placed in the hollow parts cc of Fig. 2, when the two parts were securely fixed together by the nuts and screws IG, 'G. In the same manner, the lower end of Fig. 1, was enclosed in the two iron boxes MNO, M'N'O' Fig. 3, and fas tened in that position by means of the bolts M'N', and the shears above described. The whole were then rested on the props Fig. 4, and the hook of the seals being inserted in the circular hole formed by 00' Fig. 3, the whole was ready for the experiment as shown at large in the former figure.
Every thing being thus prepared, the wedges shown in the plate were introduced under the scale to keep it steady, while the larger weights were put in. The wedges were then removed, and smaller weights ad ded in succession till the fracture took place. As a small vibration in the scale might cause a fracture in the small cylinder submitted to the operation of the weight, fou- small braces were made use of, one at each corner of the scale, to prevent any such mo tion; and every other possible precaution was had re course to in order to ensure accuracy in the results, which in this case was the more necessary, as these were afterwards to be introduced in order to examine some of the more complicated strains and resistances.
The following are the principal results obtained in these experiments, the smaller cylinders being re duced to what they would have been on square inch bars: it being assumed that the strength or resistance is proportional to the section of fracture, of which, it is presumed, no doubt can be entertained, it should be observed that these were select specimens of tim ber which had been a long time in store and perfectly dry.
These experiments were made, as is above stated, upon select specimens, and, therefore, exceed the ge neral strength of wood of their respective descrip tions; hut, on the other hand, they were made with great care, and the general uniformity in the strengths of the similar specimens shows, that we may place great reliance on the results. These strengths, in the nearest round numbers, may be stated as below.
The results, and the description of the apparatus, are taken from Barlow's Essay on the Strength of Tim ber, and they may, it is conceived, be considered per fectly satisfactory; therefore, where former results differ widely from these, they will be best omitted; for it is difficult to say whether furnishing a practical man with no information on this subject, or giving him a variety of discordant results, is most injurious? For this reason, in the following table we have omitted all those experiments given by Emerson and by Ander son, commonly found in our books: they are both clearly very inaccurate, unless they are meant to in dicate the strain that may be safely borne, and not the ultimate strength, -which probably is the case. In as certaining the direct cohesive power of metals, much less delicacy of operation is required, because they are not so liable to rupture from a want of direct applica tion of the power; but, on the other hand, they are generally made on specimens which require great force to break, and some inaccuracies are thus intro duced into the results. A remarkable case of this kind occurs in the experiments reported in the pre ceding work, made by Mr. Telford, at Brunton's ca ble manufactory, and those made at Captain Brown's manufactory; the specimens of iron were the same, yet the results differed in about the ratio of 29 to 25. Mr. Barlow has shown the discrepance is most pro bably due to the two machines: In Brunton's experi ments the machine employed was an hydraulic press, and the power exhibited by the small valve was op posed both to the friction of the piston, and to the bar, whereas it is supposed to be opposed only to the latter ; this machine,therefore, overrates the power. Captain Brown's, on the contrary, was constructed on the principle of the weigh bridges, and, conse quently, all the inertia was to be overcome before the exhibition took place in the register; we have, there fore, assumed the mean effect 27 as the truth, and have, accordingly, diminished the several results in Telford's experiments 7 per cent. and increased those of Brown's 8 per cent. They will thus be found to accord very nearly with some other experiments on which we have great reason to place reliance.