MECHANICS AND USEFUL ARTS. 83 



rials, should be correctly ascertained. Thfe portion of the investiga- 

 tion Mr. Fairbairn and his colleagues conducted by subjecting tubes 

 of different shapes to a series of experiments, the results of which 

 were briefly as follows: 1. Cylindrical tubes, on being subjected 

 to nine very severe trials, failed successively by collapsing at the top, 

 or, in other words, by evincing inability to resist compression ; 

 the tube losing its shape, gradually became elongated, or lantern- 

 jawed, while the two extremities were observed to flatten or bulge out 

 sideways, besides which, the ends, which for precaution's sake rest- 

 ed in concentric wooden beds, invariably bent inwards. 2. Elliptical 

 tubes, with thick plates riveted to the top and bottom, had been par- 

 ticularly recommended for experiment by Mr. Stephenson. These 

 tubes, under heavy pressure, displayed greater stiffness and strength than 

 round or cylindrical ones ; but, after being subjected to a variety of 

 torturing experiments of a most ingenious description, they all evinced 

 comparative weakness in the top to resist compression. They like- 

 wise exhibited considerable distortions of form. 3. A family weak- 

 ness in the head having been thus detected in all models circular at 

 bottom and top, rectangular tubes were in their turn subjected to trial. 

 As they at once appeared to indicate greater strength than either of 

 the other two forms had done, a very elaborate and interesting investi- 

 gation was pursued by Mr. Fairbairn, who, by the light of his ex- 

 periments, soon satisfied himself of the superiority of this form over 

 the other two ; and every successive test confirmed the fact. 



" The following is an abstract of the important result of about 40 

 experiments made on the comparative strength of circular, elliptical, 

 and rectangular tubes: Circular, 13; Elliptical, 15; Rectangu- 

 lar, 21. 



" As soon as the rectangular was, by the investigation, clearly as- 

 certained to be the best form of hollow tube that could be selected, 

 the next important problem to be determined by experiment was, what 

 amount of strength should be given to it, or, in other words, what 

 should be the thickness of its top and bottom, in which, as we have 

 shown, consisted its main power. The investigations on this subject 

 soon demonstrated that if, instead of obtaining this thickness by rivet- 

 ing together two or three layers of plates, they were, on the principle 

 of the beam itself, placed in horizontal strata a foot or two asunder, 

 the included holknv space being subdivided by small vertical plates, 

 into rectangular passages or flues extending along the whole top as 

 well as bottom of the tube, an immense addition of strength, with 

 very nearly the same weight of material, would be obtained. This 

 adaptation proving highly advantageous, it was deemed advisable that 

 further experiments should be made by Mr. Fairbairn and his col- 

 leagues, to determine finally the precise forms and proportions of the 

 great tubes. For this object an entirely^ new 7 model-tube, one sixth 

 of the dimensions of the intended Britannia Bridge, was very care- 

 fully constructed : and the cellular tops and bottoms thereof, as well 

 as the side?, were subjected to a series of experiments, until the 

 exact equilibrium of resistance to compression and extension, as also 

 the variations in the thickness of the plates in the several parts of the 



