Respectfully, your obedient servant, RICH'D DELAFIELD, Captain of Engineers. Brig. Gen. Charles Gratiot, Chief Engineer. * * * * * BALTIMORE, May 6, 1833. _Sir_: The instructions of the department of the 23d July last, relating to the method of repairing the Cumberland Road east of the Ohio, are founded upon principles upon which I differ in opinion, and beg leave to request your reconsideration, involving, as they do, an expenditure of not less than $250,000, when compared with what I judge to be the most judicious method of making the repairs. It is in relation to the propriety of breaking up the old bed of the road in all cases. I apprehend the department was not aware that the bed is a substantial, yet rough pavement, and not formed of loose, detached masses of quarry stone thrown together, without order. It is important to consider this particular when examining the authorities on road making. My own views are that it is decidedly preferable to retain the old pavement in all cases where its continuity is unbroken, even mending small parts that may be deranged, and Macadamizing over it. In this, I think, I am borne out by Macadam, Dean, Telford, and Farey, whose ideas on the subject are annexed, as extracted from "Macadam on Roads." The only two arguments against the method I propose are, first, that the metal will grind to dust by being placed over large stone. In answer to which, I say, that the road passing through a rocky country, even after removing the pavement, there still remains a rocky foundation; and where the pavement is well bedded in sand or clay, we have all the elasticity necessary from the clay or sand bed through the pavement. In support of which, see the sample of metal taken from the road through Uniontown, where the under strata have not worn or crushed an iota, presenting angles as sharp as the day they were first placed there. Were the metal placed upon an unyielding rock, it would doubtless soon grind to dust; but placing it upon a pavement laid in sand or gravel, preserves the elasticity so necessary for this kind of road. Second: That large stone, placed under Macadam metal, will work to the surface. This is doubtless true when detached pieces are surrounded by the metal, but with a pavement the case is very different. I find pieces of this Cumberland Road, repaired as far back as 1827, by Mr. Ewing, over the old pavement, in perfect order to this day; as, also, some parts done in this way by Giesey in 1829, that are much better than any of the repairs he made at the same time; and a piece through Uniontown, by the authorities of the place, in 1830, remains in perfect order.