compressing or extending the lead. Pipes which did not require any calix were term ed solutce* The fact referred to in the last paragraph, of the increase and diminution of the quantity of water flowing through a tube, by altering its shape, is of sufficient interest to authorise some further notice of it. It must be stated in the first place, that more water will flow through a short tube than through a simple orifice of the same diameter. It may be thus ascertained : bore a hole of an inch diameter in a bucket, plug it up, and, having filled the bucket with water, withdraw* the plug. On examination of the stream that issues from the hole, it will be found to taper off considerably at half an inch from the distance of half the diameter of the hole. If a short tube of the same diameter be inserted in the hole, the discharge of water will be greatly increased, and if at the distance of not more than two or three diameters, this tube should be made to flare gradually, or assume a conical shape, the volume of water passing would be more than doubled, as compared with the discharge throvigh the hole, without any tube. The principle upon which this is accounted for is that of capillary attraction, for if the interior of the conical tube be smeared with tallow, or any other substance which does not readily coalesce with water, the effect ceases.t This increased discharge is not confined to circular or conical tubes ; the sides of a channel may be straight and its section a triangle or square, as well as a circle. It was to guard against frauds of this nature, that the Roman Calix was adopted ; a short bent tube of brass or bronze, that formed the communication between the castellum and the leaden pipe for the supply of private houses, which pipe, by a Senate decree, was required to be of the same diameter as the calix, for the distance of fifty feet from the castellum.J The proportion in which the prodigal water of the Roman aqueducts was distri- buted, is given with great minuteness by Frontinus. The general result is, that of the aggregate supply of 14,068 quinariae, 4063 quinariee, were distributed without the city, (extra urbem^) of which 1718 were in the name of Caesar, and the residue to private use. The remaining 9955 quinariae were conveyed into the city, and received into 247 castella as well public, as private ; 3847 quinariae were appropriated to private use ; the rest went to public exhibitions, useful trades, the supply of the camps, of the amphitheatres, fountains, and pools. * Anthon's Die., art. Castellum Aquae. t Ewbank,p. 480. } Ewbank, p. 480. PRELIMINARY ESSAY. 23 Lead was universally used for the supply-pipes from the castella, and notwithstand- ing Vitruvius speaks of this metal as objectionable on the score of health, we find in Frontimis no allusion to any evil from its general employment. The liability of these gigantic works to injury and decay, especially in the portions above ground, is forcibly dwelt upon by Frontinus. The subterranean parts were easily kept in order, as they were neither subject to the action of frost, nor to the action of the summer's heat, which the Romans seem to have considered hardly less injurious to mason work. Various causes of dilapidation are enumerated by our author ; the cupidity of indi- viduals through whose lands the aqueducts passed, and who, for the sake of irrigation, or for domestic uses, were tempted to enlarge any holes through which the water (Aqua Caduca) oozed out ; the violence of tempests, the imperfection of the work, especially in the more modern structures, and the softness of the material, tuff a, too frequently em- ployed in portions of the mason work, where there was great pressure. The arcades, and especially those which traversed streams, were particularly liable to damage from the violence of storms. Another considerable source of repair, arose from the adhesion of the sediment to the sides and bottom of the water-channels, forming a thick, hard crust, which materially ob- structed the passage of the water, and by eventually raising its level, occasioned breaks in the channel, whence the waters escaping, not only destroyed it, but the sub-structures of every kind in the vicinity.* All work of repair, was however, as much as possible suspended in the summer sea- son, as then the free use of the water was most needed and agreeable.t The spring and au- tumn were the working periods. Moreover, a moderate temperature was deemed advan- tageous, as permitting the masonry to be laid with the degree of humidity deemed essen- tial to its perfection and ultimate solidity, excessive heat and