OF CYLINDRICAL VAULTS.
9
resist compression, and commenced a course of experiments to de-termine the strength of stone in that respect, in conjunction withSoufflot; which was subsequently pursued by Gauthey and Ron-delet. Though a very daring architect he did not venture to con-sider a practical strength at more than one-twelfth of a breakingstrength. In the investigation of pendent bridges the author hastaken one-fourth of the breaking strength for his practical strengthThe principles upon which the thickness of a chain is determinedin that investigation are applicable to the solution of the problem“ to determine the height of the key-stone of a given arch.” Thespace or opening has been the general measure of its height, andarbitrary rules have been given varying from l-15th to l-30th.Belidor was the first to distinguish between strong and weak stone ;and Blondel between semi-circular arches and arches surbaissees.Tables and rules are given in the works of Belidor , Gautier, Blon-del, Gauthey, and Rondelet. Gauthey says of the rule which hegives, that it is one “ purement empiriqueand the same conclusionmay be come to in respect to the others. The tables in the 6thvol. of Patte’s edit, of Blondel’s Cours d’Archi, p. 197, are drawnup very aptly for practical purposes, if reliance can be placed uponthe formula by which they were calculated. The French archi-tects have elicited, that the radius of the circle of curvature at thevertex is the proper measure of the height of the key-stone: andin order that this might be a given quantity, and that the directionsof the joints may radiate to points, they have adopted, like Mr.Milne at Blackfriar’s-bridge, the deformed curve called Anse depanier, or false ellipse, for the curves of their arches surbaissees.Perhaps the example set by Mr. Rennie at Waterloo-bridge maybring the conic section into favour in France . It has been shewn,in the inquiry respecting pendent bridges, that a sheet of flexibleiron an inch thick, calculating the practical strength only at one-fourth of the breaking strength, and, supposing there to be no in-cumbrance, might be suspended in the curve of a catenary over a