li E IMJ R T UPON THE THIRD SUBDIVISION OF THB CENTRAL TRANSPORTATION I ROUTE, FDOAI THE OHIO OR KANAWHA RIVER TO TIDE-WATER IN VIRGINIA, IN OMAHGK OF WM. P. ORA^IOHILL MA.TOK OF ENGINEERS, UVT. LIEFT. COLONEE, U. S. A.; BEING APPENDIX V OF THE ANNUAL REPORT OP THE CHIEF OP ENGINEERS FOR 1877. WASHINGTON: aoVKBNMBNT FEINTING OFFICE. 1877. \X^ E'rv^\Y\ee<^ , EE POET UPON THE THIRD SUBDIVISION OF THE CENTRAL TRANSPORTATION ROUTE, FROAI THE OHIO OR KANAWHA RIVER TO. TIDE-WATER IN VIRGINIA, IN CHARGE OF WM. I>. CR^AiamiLlL, MAJOR OF ENGINEERS, BVT. LIEUT. COLONEL, U. B. A.; BEING APPENDIX V OF THE ANNUAL REPORT OF THE CHIEF OF ENGINEERS FOR 1877 WASHIJIGTONT: GOVERNMENT PRINTING OFFICE. 1877 . ' "n ■ ['-'1 '( .1 ' ■ ) \ E V ■:? \ I ^ 'i*r; rC)i:j;r^ H;:: '^1 3 ^ / 'i u a 0 4) Ul UT.\ [EXTRACT FROM THE ANNUAL REPORT OF THE CHIEF OF ENGINEERS TO THE SECRETARY OF WAR.] Office of the Chief of Engineers, Washington^ D. C., October 12, 1877. TRANSPORTATION-ROUTES TO THE SEABOARD. The surveys of the third subdivision of the central route, designated as — A connection by canal or a freight-railway from the Ohio or Kanawha River, near Charleston, by the shortest and most practicable route through West Virginia to tide- water in Virginia — were made, in compliance with the provisions of the act of June 23, 1874, under the supervision of Maj. William P. Craighill, Corps of Engi- neers, who submitted a final report, dated November 10, 1876. This report was printed as Executive Document No. 15, Senate, Forty-fourth Congress, second session.' (See also Appendix Y.) #**#*#* O r~ APPENDIX V. REPORTS ON TRANSPORTATION-ROUTES TO THE SEA- BOARD. THIRD DIVISION OF THE CENTRAL TRANSPORTATION-ROUTE. United States Engineer Office, Baltimore, Md., Novemher 10, 1876. General : Early in June, 1874, instructions were received from you, from which the following extract is made : The river and harbor act, approved June 23, 1874, contains an appropriation for sur- veys and estimates for the improvements recommended by the Senate Committee on Transportation-Routes to the Seaboard upon four routes indicated in the report of said committee, to be expended in such manner as will secure the greatest amount of exact information for each of said routes. The survey of that portion of the central route designated as ‘^a connection by canal, or a freight-railway, from the Ohio River or Kanahwa River, near Charleston, by the shortest and most practicable route through West Virginia, to tide-water in Virginia,” is assigned to you. A prelimiuary report was submitted January 13, 1875. Reports on the subject of the proposed freight-railway will be found therewith from Mr. H. D. Whitcomb and Mr. C. P. Manning. No additional informa- tion has been since procured as to the freight-railway. The surveys for the water-line were made in 1874, under the personal supervision of Lieut. Thomas Turtle, Corps of Engineers, and Mr. N. H. Hutton, assistant engineer. No reports from these gentlemen accom- panied the report of January 13, 1875, for reasons therein stated. Since that time the preparation of maps, estimates, &c., has been carried on in connection with the current labors of this office. In these labors Lieutenant Maguire, Corps of Engineers, assisted Lieutenant Turtle most zealously until his detail to the staff of General Terry, for service in the Indian country. It is not considered necessary to introduce here the names of all the gentlemen who assisted so efficiently in the surveys and in the office, as they are placed upon the maps. Quite full reports from Lieutenant Turtle and Mr. Hutton are hereto appended, accompanied by estimates in detail and illustrative maps and other drawings. These relate to the summit division, the Greenbrier division, and the New River division. A location for the long tunnel at the summit was made by Mr. Will- iam R. Hutton in 1870. When this subject was under consideration by the Board of Engineers of 1874, of wliich Bvt. Maj. Gen. J. G. Bar- nard, colonel of engineers, was president, he suggested the examination of a tunnel-line from Brush Creek to Howard’s Creek, or to the Green- brier River, for reasons stated bv him. Surveys of both these lines w^ere made under Lieutenant Turtle’s immediate supervision. His inter- esting report and maps indicate the details of the surveys and the re- sults. He discusses quite fully the cross-section, interior arrangement, ven- tilation, &c., of the tunnel, the methods of towing by animal-power or by steam, the use of a chain or cable in towing, the kind of fuel, the passage of boats singly or in fleets, the elevation of the tunnel above 676 REPORT OF THE CHIEF OF ENGINEERS. tide, the advantages of the several locations as to shafting, &c., the debouches, the rate of execution, the arrangements for feeding, &c. Surveys were also made, under Lieutenant Turtle’s supervision, down the Greenbrier River, resulting in locations and estimates for a slack- water navigation or an independent canal. The survey of the New River division was made under the personal direction of Mr. N. H. Hutton, with a view to a location and estimates for slack-water navigation, as well as for an independent canal. The * shortening of the line by cut-off tunnels was also considered. For the cost of the enlargement of the James River Canal from Rich- mond to its present terminus, (Buchanan,) and its proposed extension thence to the mouth of Fork Run, according to the location of Mr. Lor- raine, reliance is still placed upon the estimates qf Mr. W. G. Turpin, an abstract of which will be found in the report on the water-line of Jan- uary 27, 187 L The detailed estimates of Mr. Turpin are appended hereto. In the report of January 13, 1875, brief mention was made of the Great Kanawha River as a part of the central water-line. In March, 1875, there was an appropriation by Congress of 8300,000 for its im- provement, and a second one of $270,000, August 14, 1876. A special report on that improvement, by the superintending engineer, dated April 30, 1875, and a report of a Board of Engineers, dated May 25, 1875, may be found in part 2 of the Annual Report of the Chief of Engineers for 1875, beginning at page 89. The work contracted for under the appropriations mentioned above, is in pursuance of the plan of improvement recommended by the Board, viz, of large locks, with movable dams, from the mouth of the river to Paint Creek, and permanent dams at and above that point. A good map of the Kanawha has been in existence for some years, made by Mr. Byers, from his own surveys in 1856. Other examinations have been made of portions of the river, by Ellet, Gill, Lorraine, and others. In 1873 and 1874, special additional surveys were made by Mr. A. M. Scott, whose report is herewith, accompanied by estimates. A special report by Mr. William R. Hutton is also appended. The following is a summary of the estimates taken from the appended reports : Mr. Turpin’s estimate for enlarging the existing James River Canal as far as Buchanan, its present terminus, and for constructing extension thence to the mouth of Fork Run, according to the location of Mr. Lorraine, is. $14,781, 000 Deduct cost of lock and ship-lock at Richmond 1, 300, 000 Total from Richmond to east end summit division 13, 481, 000 Summit division, Lieutenant Turtle’s estimate 16, 387, 000 Anthony’s Creek reservoir, Lorraine’s estimate 300, 000 Total from Richmond to '^est end of summit division 30, 168, 000 For the slack-water project, Greenbrier division, Lieutenant Turtle 6, 251, 000 New River, Mr. Hutton 11, 427, 000 Removing bowlders in New River, Mr. Harris'’^ 260, 000 Kanawha division 4, 000, 000 Total 52, 106, 000 ^ The total estimate of Mr. Harris for removing bowlders from the bed of New River is $307, 385 From this, in order to know the proper sum to be added to Mr. Hutton’s slack- water estimate, must be deducted the amount estimated as necessary for clearing the river in places where the line is not located in the open river. This amount is 46,875 Total 260, 510 APPENDIX V. 677 For tbe independent canal : From Richmond to west end of summit division, as before $30, 168, 000 Greenbrier divisioy, Lieutenant Turtle - 4, 765, 000 New River division, Mr. Hutton . 20, 650, 000 Kanawha division 4, 000, 000 Total 59, 583, 000 The estimates are very full, and it may be confidently expected that the cost of the work, if executed, will be less than the estimate, if money be provided as fast as needed for economical progress.* Should this water-line ever be opened, it would doubtless only be after a careful revision of the whole subject by a board of engineers. It seems superfluous, therefore, to do more now than put on record a few general statements. The careful surveys and estimates made since the report of the board of engineers, dated March 18, 1874, prove the correctness of the opinion expressed in the following resolution, unanimously^ agree;l to : Resolved, That, in the opinion of this board, it is entirely practicable to connect th& waters of the James and Ohio Rivers by a water-navigation of seven feet depth. It will be interesting also to recur to another resolution of the same board, which was agreed to by four of its five members : Resolved, That, in the opinion of this Board, the water-line by the James River and Kanawha route, with seven feet depth, may be completed in six years at a cost of not more than $60,000,000, allowing an unusually broad margin for contingencies which cannot be accurately measured. The cost may be reasonably expected to be within $55,000,000, and possibly will not exceed $50,000,000. The existence of the Chesapeake and Ohio Eailroad has added greatly to the estimated cost of the water-line. The building of another railroad along the New and Greenbrier Eivers would probably make the cost of the water-line so great as to be prohibitory of its construction. As attention has been directed to the line of the Chesapeake and Ohio Canal, as a proposed substitute for the central water-line by way of the James, Greenbrier, New, and Kanawha Eivers, it seemed proper to compare the two as to their respective features. Accordingly Lieu- tenant Turtle has, at my request, prepared a comparative statement, which is appended hereto. Objection having been made by some parties to the proposed improve- ment of the Ohio Eiver by locks and dams, and the same having nearly as much pertinence to the similar improvements of the Great Kanawha Eiver, a forcible reply thereto (devoid of professional technicalities and therefore suited to the people generally) is appended hereto, which ap- peared in the columns of the Pittsburgh Commercial. The original report of McNeil, of 1828, on the water-line, must always be of great interest and value in the study of the subject of which it treats. As it is nearly unobtainable at present, a copy is appended to this report, in the hope that it may thus be put again in print. For the same reason are added copies of special reports on the Great Kanawha, by Mr. John A. Byers, dated February 1 and 10, 1868, and copies of re- ports in January and October, 1852, by Mr. E. Lorraine, on his survey of the summit-level. The United States having begun the improvement of the Great Kana- wha Eiver by locks and dams, it should not be forgotten that certain rights and privileges have been granted, under the laws of Virginia and West Virginia, to a board or company, relative to the improvement of the ^ The details of the estimates are omitted, but are to be found in the records of the Engineer Bureau, with the maps and note-books relating to this subject. 678 REPORT OF THE CHIEF OF ENGINEERS. Kanawlia, tlie collectiou of tolls on navigation, &c. Copies of these laws, &c., as far as known to me, are herewith. In this connection, the enactment is suggested for airplication to the Kanawha of a law similar in. its provisions to the act of Congress ap- proved March 3, 1875, to aid in the improvement of the Fox and Wis- consin Rivers in the State of Wisconsin,” of which a copy is herewith. Very much has been written concerning the central water-line. Those who wish to study the subject are advised to consult, in addition to the papers appended hereto, the following, which are in print : 1. Report dated January 27, 1871, and attached papers ; (see Ex. Doc. No. 110, House of Representatives, Forty-first Congress, third session ;) also, printed in part in Annual Report of Chief of Engineers, 1871, begin- ning at page 624. 2. Reports, dated December 12, 1872, and April 11, 1873. (See Annual Report of Chief of Engineers, 1873, beginning at page 828.) 3. Report dated March 18, 1874. (See Annual Report of Chief of En- gineers, 1874, part 2, beginning at page 86.) 4. Report of the select Committee on Transportation-Routes to the Seaboard, with appendix and evidence, being Senate Report 307, parts 1, 2, and 3, Forty-third Congress, first session., 5. From page 87 to 98, part 2, Annual Report of Chief of Engineers, 1875. 6. Report of January 13, 1875, printed in Senate Ex. Doc. 19, part 2, Forty-third Congress, second session, and in the Annual Report of the Chief of Engineers for 1875, part 2, beginning at page 631. 7. Ellet^s report on the Great Kanawha, 1858. 8. Annual Report of the Chief of Engineers for 1876. Repectfully submitted. Wm. P. Craighill, Major of Engineers. Brig. Gen. A. A. Hu3IPHREYS, Chief of Engineers^ JJ. S. A. COMPARATIVE STATEMENT OF DISTANCES, ETC., RELATIVE TO THE CENTRAL WATER- LINE AND THE CHESAPEAKE AND OHIO CANAL, BY LIEUTENANT THOMAS TURTLE, CORPS OF ENGINEERS. Baltimore, Md., August 16, 1876. Major : According to yonr request, I submit the following comparative statement of distances, &c., relative to the central water-line and the Chesapeake and Ohio Canal. The horizontal distances are from the following authorities : Richmond to City Point. — Coast Survey map. City Point to Newport News. ^ Newport News to Capes. | Newjjort News to Georgetown, y State map of Virginia. Newport News to Baltimore. | Georgetown to Baltimore. J Point Pleasant to Richmond. — From the various surveys and reports of the James River and Kanawha Canal ; of the central water-line, and reports on the Kanawha. Pittsburgh to Georgetown. — Reports on Chesapeake and Ohio Canal. Pittsburgh to Cairo. — Reports on Ohio River. Cairo to Saint Paul. — Reports on Mississippi River, and report of Senate Committee on Transportation. Cleveland to Portsmouth. — From report of Senate Committee on Transportation. Table No. l.—Showing horizontal distances between certain points via central water-line and via Chesapeake and Ohio Canal. AIPENDIX V. 679 •|BnBO TrnB -TJ^ BIA pOB^AO^O 1 iO 1 ^ 0 10 ^ 0 T?^ 0 lO 10 CC 1-^ CC Cl •H 0 1139.5 1095. 5 lO < 2 i . 10 I 05 0 I 1 0 rc » • 05 CO • 05 CO CO 0 cc •oiqo 977.6 10 CO rH ^ C30 00 Oi CO 10 lO 00 0 CO CO CO CO lO CO ^ GO Cl Cl CO GO 10 i t- » ....... 514. 4 563 05 GO • A\ ‘ganqeaaj^j'B,! 968.5 CO 0 lO 0 t- 0 CO lO 10 tA CJ lO CO CO 0 AO § 2 CO 00 lO i g : 523.5 554 CO 00 0 00 •a9AT^ Sion -^11 uiA ogBoiqo 0 01 2311 1836.1 2018 1811.5 0 d> TJ- i- 0 05 d 1-1 CO * d I • 10 I • CO , • GO r-* • lO 0 1C 0 d •nniH ‘itiB j 7niBS 2530. 6 2654. 6 2237. 6 2419. 6 2153 2445. 5 2381 S : ^ ; • 10 » IfS Ci . 0 r • Cl • Cl G 3 I -i • 1925 d CO tH ■sjAi ‘nOITIO tip 9TJTBJ t 540.5 i XO • . 0 • • '•r r:i • •CO r- * •BA ‘pnoniqDTH 10 0 5J 0 C5 lO 10 d i 1 ir • 1 ^ 1 • t' • 1 c : • i I ’0 -a; ‘uAiojaSjoaf) t- -0 0 GO 0 ■pn ‘Ovtouii^iBa lO Cl CO 00 CO •8M.9j!l !|J0d.U.9j^ j t-- Cl •BmigjiA JO S9dB3 1 ^5 a Oi •A!qo qaoA .S4.9JSI; j : eS 03 -fj a o OD © p CQ 2 © CT ® © 73 -(J 'w’E* .20 § atSo .251 2 p a « ^ .s +=0 © .a 'S c.i> ® © p 43 0.0 ®.S ® P» 3 ® .2 %as 5 § 2 i . b> P O ^ OPS 680 REPORT OF THE CHIEF OF ENGINEERS. Table No. 2 shows the total lift (ascending and descending) between certain points. Table No. 2. Feet. Pittsburgh to Georgetown via ChesaiDeake and Ohio Canal, about 3, 187 Point Pleasant to Pittsburgh 495 Parkersburg to Pittsburgh 138.6 Marrietta to Pittsburgh 13*2. 2 Point Pleasant to Richmond 2,911 Point Pleasant to Parkersburg 48. 5 Point Pleasant to Marietta 55.3 Table No. 3 shows the equated distances between those points affected by the two lines, (Central, and Chesapeake and Ohio.) The Central is intended, at the outset, to be operated by steam. The experiences of the William Baxter steam canal-boat on the Erie Canal indicate that with steam one minute to each foot of lift is an ample allowance of time in equating distance, and such boats, when freed from the petty and unjustifiable annoyances and hinderances put upon them by horse-boats, will be able to make three miles an hour when fully under way, without injury to the banks. I estimate that the summit-tunnel of the Central line (single width, with turnouts for fleets) will cause an average delay of 19^ miles — say 20 miles. Colonel Sedgwick estimates that the use of the proposed planes on the Chesapeake and Ohio will save 52 miles in equated distance. On the basis supplied by the above, the equated distances of Table 3 are compiled. The horizontal distances corrected for lift are increased by 20 miles for the Central line and diminished by 50 miles for the Chesapeake and Ohio, the 2 miles additional saved by the planes being supposed to be lost in the summit-level. Table No. 3. — Showing equated distances hetiveen certain points via central watci'-line and via Chesapea'ke and Ohio Canal. APPENDIX V. 681 BIA paB[9A8{0 ii s 1267. 5 1083 127.3. 5 1251 1293. 5 1093. 5 985.6 in i •otqo 00 00 i! s T}< 00 i = rp X I i i i i CO i X rH- CO "KA 'jA ‘SJuqsiaqjBj 1126.4 1106 8.33. 4 2 § X X X i s X i X i g 1 •j;9Ai^3; stounni BiA oSBDtqo 0 10 ii i i Ol § ^ o? 55 2 I 1 i to I •nntiY ‘[UB j luiBg 0 ro X •n* Ci ox cc Cl Cl Cl CO a to CO to CO Cl Cl to C5 to CO to to 0? Cl to c: t- ci to CO CO rr* Cl Cl to CO Cl Cl cd ■stAi ‘naiqo up emBij 10 0 CO X 0 — CICK CM to cd CO 3! to d d to cd rfi c? CO CO Cl 0 CO Cl CO CO cc ^ 0 Cl Cl to cd §5 •III ‘pu^isi qooH to 00 0 lO 0 Cl ^ 0 d Cl Cl TT 0 CO X — C: Oi ^ to 0 X Cl Cl to CO Cl CO § s CO X rH T— 1 to 0 to CO to Cl •J9AtJJ SToaini jo' qjnoj^ to d c; — ^ cd C5 CO •T X CO to ci 1 - CO rH X X to to X to CO Co rH to to CO § Cl Cl c- "OR ‘eiiioq; jniBg LO 10 CO Cl C? X CO Sg s to oi QL C5 0 i- CO X CO C5 p- t- to x' X to CO 22 ® 0 to cd CO X X •III to cd c? CJ X CO X to d X ^ 0 to X CO a to to to GO to CO 00 0 CO to 00 CO CO X CO •A3 ‘ani^-sinoT; m C5 0 CO i-C d CO CO iO 0 to cd Cl Cl T?" 0 ? 0 to Cl 0 d ^ to ci to Cl 10 CO Cl -r to 0 CO to Cl 0 CO to CO •oiqo ‘ijBaapntf) 10 X dOD CO OJ CO 0 0 s to 0 X 0 1 - r-i 0 i to CO 0 Cl Cl rH CO X to 0 CO X •oiqo ‘qjnonisiJOti to to CO 05 X t- 0 rH 0 X to to 0 X 0 05 to Cl' 0 to CO t-‘ CO* Cl rH X «> to to CO 0 00 •BA •AV ‘lUBStsaij jnioj to to ^ - 9 * 0 to 0 CO i> to rH l> rH OJ c: i> to to rH 0 CO X to CO 0 to CO t-* ci CO Cl i> CO to t--* X Cl •BJ ‘qgjnqs^jx^ to X to rH cd X CO S 0 X to Cl — CO 0 X to CO cd d X CO rH X CO CO X CO t- TT CO Names of points. 0 1 ! & i ! 1 • fi : ts 1 ^ ! ^ 1 1 f i ? g p > S' rj t g 5 682 REPORT OP THE CHIEF OF ENGINEERS. From this table the following deductions are drawn : If the capes of Virginia he the objective point, the shortest line for all points west of Point Pleasant will be via cen- tral w^ater-line. Difference in favor of central, 160 miles. Same if New York be the objective point. If a port on tide-water (Georgetown for Chesapeake and Ohio or Rich- mond for central) be the objective, 108 miles is saved by the central for all points west of Point Pleasant. If Baltimore be the objective, 42 miles is saved to ail points west of Point Pleasant by the central. If a port on Hampton Roads (say Newj5ort News) be the objective, the distance saved to all points west of Point Pleasant by the central is 190 miles. On the central water-line it may be advisable to adopt one or two double-track planes (or a hydraulic lift) at the eastern approach to the tunnel, instead of the two flights of locks. (Vertical distance to be overcome, 64 feet.) According to Major Mer- rill’s report, the plane at Georgetown, overcoming a lift of 36 feet, will cost about $100,000. The boats on the central will be much larger than those in present use on the Chesapeake and Ohio Canal. If we suppose two planes at the eastern approach to the summit-tunnel of the central to cost $200,000, the saving in estimated cost would be $126,308.57. A plane or vertical lift might be used at Richmond instead of the flight of locks. On the Greenbrier division it might be advisable to place one plane below Alderson, changing the location as at present made to the following: Leaving dam No. 11 with canal-bottom at the elevation 1,602, as at present, and keeping this elevation till the head of the flats opposite Alderson is reached, and then by a plane overcome the lift of 52 feet, six locks would be saved. I presume the excavation would be about the same as now estimated by including that for the locks. The six locks, exclusive of excavation, would cost about $144,000. Colonel Sedgwick estimates the cost of a plane of 64 feet lift at $43,331.75, which Major Merrill doubles, on account of the experience at Georgetown. It maj" be, then, that the plane of 52 feet lift would cost one-half as much as the locks, and a saving of $72,000 would be made. These locks are single, and will have to be doubled, without doubt, w'hen the trade of the line is developed. It may be that a jAane would be advisable in the line round the Great Bend, but our data are not sufficient to permit a recommendation to be made. If the higher summit (McNeill’s) should be adopted, it might be that several planes would be advisable in overcoming the lift from the tunnel to the mouth of Fork Run, (about 288 feet.) I doubt if the use of planes would be advisable between the western portal of this tunnel and the mouth of Howard’s Creek, on account of the quite gradual fall (compared with Fork Run Valley) between these points. The total lift to the first pool on the Greenbrier from McNeill’s tunnel is 246 feet. The distance from the west- ern extremity of the approach cut to the mouth of Howard’s Creek is about 45,000 feet. This would require that the locks (of 8 feet lift) should be at the distance of about 1,400 feet apart, (average.) It might be found difficult in the location of planes to keep the canal upon the hill-side to gain sufficient height. The tunnel here proposed by McNeill would be, according to his report, 13,920 feet in length. This is with a depth of cutting of 50 feet. The jmoposed tunnel on the summit-level of the Chesapeake and Ohio Canal is 19,800 feet in length. By a slight variation from McNeill’s location, and increasing the depth of approach-cuts to 80 feet, the length of the tunnel could probably be reduced to 12,000 feet or less. While referring to this higher summit, I would call attention to the consideration that the short line might permit us to excavate two parallel tunnels at the outset. Then, as the slowness of movement w'ould not interfere with the carrying capacity of the line, the tunnels might each be made of less section than would be permissible wTth a sin- gle tunnel. McNeill’s summit on the central water-line is 1,916 feet above tide. The proposed summit for the Chesapeake and Ohio Canal is 1,944 feet above tide. Th^ summit of the central w'ater-line is about 2° 10' south of that of the Chesapeake and Ohio. The ijresent terminus of the James River and Kanawha Canal is 812 feet above tide. That of the Chesapeake and Ohio is about 624 feet. On the James River and Kanawha Canal, from 1848 to 1868, the number of days in which the canal was closed by ice varied from none to fifty-six, and for the twenty years the average was 15.1 days. The Chesapeake and Ohio Canal closes from 1st to 15th of December ; opens about 20th March. On the central water-line, as now estimated for, the canal-trunk is at no point, except at aqueducts, of less width than 56 feet on the bottom. A portion of the Chesapeake and Ohio is but 45 feet. The proposed canal-locks on the central water-line are 24 feet wide. Those for the Chesax>eake and Ohio are but 20 feet wide. This makes little difference in the esti- mates for the locks, but it makes considerable difference in the question of water-sup- yfly and width of the tunnel, and some difference in the cost of y^lanes and the machin- ery thereof. APPENDIX V. fi83 The nse of steam on the canal will decrease the detention at locks, and will decrease the superiority of inclined planes and the equated saving of distance due to their use. Respectfully submitted. Thomas Turtle, First Lieut, of Engineeis. Maj. William P. Ckaighill, Corps of Engineers, U. S. A. LIST OF MAPS, ETC., ACCOMPANYING REPORTS PERTAINING TO THE SUMMIT AND GREEN- BRIER DIVISIONS. Baltimore, Md., Julg 27, 1876. Major : The following is a list of the maps, &c., pertaining to the summit and Greenbrier divisions of the central water-line, and accompanying the reports of the survey of 1874 : SUMAIIT DIVISION. 1 map (sheet No. 1) of survey from Dunlap’s Creek to the Greenbrier division, scale 1 ^' = 1 , 000 '. 1 map, (sheet No. 2,) scale, 1" =200', of Brush Creek Valley. 1 map, (sheet No. 3,) scale, 1" = 200', of Howard’s Creek Valley. 1 map, (sheet No. 4,) scale, 1" = 200', of feeder-line. 38 sheets of cross-sections of Brush Creek Valley. 42 sheets of cross-sections of Howard’s Creek Valley. 3 sheets of cross-sections on opposite hill from the northern line, (ravine on the Greenbrier.; 23 sheets of cross-sections of feeder-line. 1 profile of feeder-dam. 3 transit note-books. 7 note-books of cross-sections. 18 level note-books. GREENBRIER DIVISION. 3 maps, scale, 1" = 200', of Greenbrier Valley, from the feeder-dam to New River. 1 map, scale, 1" = 600', showing the entire (^reat Bend. 4 sheets of cross-sections at Bacon’s Falls. 1 sheet of cross-sections of bar at Alderson. 37 sheets of profiles of dams and dikes. 30 sheets of cross-sections of lock-sites for slack-water. 6 sheets of profiles of dams for canal-lines. 193 sheets of cross-sections of main canal-line. 1 sheet of cross-sections of line on left bank below Alderson, (station 215 to station 22.5,) pertaining to canal-line, from dam 18 to pool at Wolf Creek. 37 sheets of cross-sections from west iDortal of Great Bend tunnel, Chesapeake and Ohio Railroad, to the mouth of the Greenbrier River. 1 sheet of cross-sections of lines on the right bank above Alderson, (approach to pool of dam 16,) pertaining to canal-line on the left bank. 1 sheet of cross-sections above dam 16 on left bank. 63 sheets.of cross-sections’of left bank below Alderson, (dam 16 to the Great Bend.) 7 transit note-books. 1 note-book of soundings. 1 profile-level book. 1 miscellaneous-level note-book. 5 level note-books, profiles of dams and cross-sections of lock-sites. 11 note-books, levels on slack-water and canal-line surveys. 17 level note-books, cross-section of canal-line. Very respectfully, your obedient servant, Maj. William P. Craighill, Corps of Engineers, U. S. A. Thomas Turtle, First Lieut, of Engineers. 684 REPORT OF THE CHIEF OF ENGINEERS. EEPORT ON LOCATION OF TUNNEL BY LIEUTENANT THOMAS TUETLE, COEPS OF EN- GINEEES. Baltimoee, Md., July 5, 1876. Majoe : I have the honor to submit the following report of that portion of the sur- vey for the “ central water-line” which was placed in my charge by your letter of July 9, 1874. My instructions, as contained in this letter, were, in part, as follows: You will proceed to the neighborhood of the Lorraine tunnel, on the Allegheny sum- mit of the central water-line, and there undertake such further investigations and surveys as may be required to furnish the information needed, to enable a definite and final location to be made of that important feature of the line, and to put the work promptly under contract should Congress provide the means. While locating the tunnel, as a means of passing a great communication through the mountain, you will bear in mind also its office as the summit-level of a canal, and consider carefully the best means of connecting it with the canal or slack-water at either end, and of main- taining its supply of water by suitable feeding arrangements, assuming that supply to be sufficient.” In accordance with these instructions, parties were formed about the middle of July, and work commenced on the 20th of that month. The transits Avere in charge of Mr. R. H. Talcott and Mr. S. F. Adams, and the levels were taken by Messrs. J. A. Harris, Henry Fairfax, Thomas Bernard, and Marsden S. Manson. I take pleasure in acknowl- edging the zeal shown by all these gentlemen throughout the survey, and their faith- ful performance of whatever duty was required of them. It was understood that the primitive object of the surveys in the vicinity of the summit was to determine the ad- Ausability of adopting a tunnel-line from Brush Creek (otherwise Jerry’s Run) to Howard’s Creek, or to the Greenbrier River direct, as a substitute for the line proposed by Mr. W. R. Hutton, in his report to you after his survey in 1870, printed in the Re- l)ort of the Chief of Engineers for 1871. This substitute was suggested, subject to the test of actual survey, by Bvt. Maj. Gen. J. G. Barnard, colonel Corps of Engineers, to the Board of Engineers on the James River and Kanawha Canal, convened by Special Orders No. 17, War Department, Ad- jutant-General’s Office, January 27, 1874. This suggestion was incorporated in the re- l)ort of the board to the Chief of Engineers, dated March 18, 1874, in words as follows : “It is also suggested that the tunnel and canal construction may he improved by a radical change of location, taking a point near the railroad-crossing in the ravine of Brush Creek (or Jerry’s Run) for the eastern terminus, and the same point on Howard’s Creek for the western. By this it is supposed that 2 miles of canalling would be saved, and the location laid in a more open valley (Brush Creek) than Fork Run ; or finally, and possibly, by starting from the last-named point and tunneling a distance scarcely exceeding that originally designed by Mr. Lorraine, (9 miles and a fraction,) the valley of the Greenbrier may be reached, by which the expensive canalling in Howard’s Creek and the feeder would be wholly dispensed with. The modifications of location are not mentioned as matters of x>ositive recommendation, but as subjects for further survey, with a view of having the best possible location.” The survey began at a point near the junction of Brush and Dunlap’s Creeks. A line was run up Dunlap’s Creek, to connect with the initial point and bench-mark of Mr. Hutton’s survey of 1870. This was necessary, as Mr. Lorraine’s line and benches of 1853 had disappeared. The surveys extended up Brush Creek to and above the rail- road fill. Cross-sections were taken of the valley, at distances of 100 feet, and extended on either side as far as was judged necessary to enable a canal location to be made. An offset was run from station 11 of the line to Dunlai)’s Creek, and cross-sections were taken on the offset ; which enables us to make a connection with Lorraine’s location in the valley of Dunlap’s Creek and in the pool of dam No. 8 (station 745) of that loca- tion. This completed the examination for the eastern approach and connection. Station 120 -j- 15 of the Brush Creek line was taken as the initial point for the trial lines for the tunnel location. The directions were obtained as nearly as possible, for the objects in view, from a copy of Captain McNeil’s map and from a plot of the Ches- apeake and Ohio Railroad, kindly furnished us from the engineer office of the company. The line from Brush Creek to Howard’s Creek is designated on the map “ Southern line of 1874,” and that from Brush Creek to the Greenbrier River as the “ Northern line of 1874.” The surveys of, and pertaining to, the southern line were made by Mr. Tal- cott, and those for the northern line by Mr. Adams. All surveyed lines were run with transit instruments. Those depressions, which offered any chances for the location of shafts, were surveyed to the right and left of the main lines, as far as was considered necessary for such locations. The lines of levels were carried along the main lines and all surveyed lines. The southern line surveys were continued down Howard’s Creek to the Greenbrier River, and cross-sections were taken in the valley of the former down to a point below Caldwell Station of the Ches- apeake and Ohio Railroad, (to station 108 of the Howard’s Creek line.) The surveys for the northern line extended to the Greenbrier RiA'er and up the river for 7,532 feet. An offset line was run up the valley of the middle fork of Howard’s APPENDIX V. 685 Creek, and another down Howard’s Creek and up a ravine, beading toward the Green- brier. The necessary surveys were also made for feeding arrangements from the Greenbrier River. The information resulting from these surveys has been compiled, and is shown in a series of four maps accompanying this report. Sheet No. 1, on a scale of one inch to one thousand feet, shows the entire ground surveyed from Dunlap’s Creek to Greenbrie River. The contours show the exact elevations where they cross surveyed lines, and are but approximate of these lines, except in Brush Creek and Howard’s Creek Valleys and along the line of the feeder, cross-sections having been taken in these localities. Upon this map are also shown Hutton’s line and i^roposed location of 1870, and Mc- Neil’s proposed tunnel of 1826. In addition to the surveys mentioned above, the strike and dip of the rock were noted in many places where it outcropped upon the lines ; and an examination was made by Mr. Talcott and Mr. Manson of the cuttings and tunnels of the Chesapeake and Ohio Railroad from Brush Creek to Hart’s Run. The strikes and dips of the different strata were noted, and specimens (126 in number) of all the different varieties of the rock were secured. The location of the tunnel-line for the summit-level is the first matter to be definitely fixed, as the connections at both ends depend upon it. It is assumed that the elevation of the summit shall be somewhat higher than the level of the Greenbrier at its western end or approach. I find it advisable to first consider the dimensions and form of the cross-section of the tunnel, and that the determination of this matter involves incidentally the discus- sion of all matters pertaining to the summit division. The late Board of Engineers were “ unanimously of the opinion ” (see report of Chief of Engineers 1874, Part 2, page 90) “that a tunnel of the dimensions proposed, (54 feet broad by 34 feet high,) i. e., wide enough for passing everywhere, should not be attempted, and, on the sugges- tion of one of its members,” united “ on the recommendation of a single tunnel with turnouts of dimensions ” to be “ fully set forth in his individual report, with which hereafter, if found necessary^ a second tunnel might be combined.” The tunnel so sug- gested was to have a water-way 34 feet wide and 7 feet deep, and would have passing- places 140 feet in length at every fourth of a mile. These passing-places would divide the tunnel into a certain number of compartments, each of which could he occupied hy hut one boat at a time, unless the length of the passing-places be increased to permit more than one boat to enter. In such a tunnel, the canal being worked regularly and to its full capacity, system would require that the boats moving in the same direction should occupy alternate compartments, the remaining compartments being occupied by boats moving in the opposite direction. A direct relation exists between the speed of the boats (taken in connection with the distance between turnouts) and the number of boats which can pass out of the tunnel in a given time. If we suppose t = fraction of an hour lost at each turnout, w=rnumber of turnouts per mile, and refraction of an hour occupied in the passage of 1 mile unobstructed, then will fraction of an hour lost at turnouts per mile, and r-f-wt— fraction of an hour required to move 1 mile, including delays at turnouts. In the time {v-\-nt), n boats will have emerged from the tunnel. Representing by N the number of boats leaving per hour, we have ?i=N (v-\-nt). The tunnel will be an enormous obstacle in the line of communication, unless it per- mits the passage of boats as rapidly as they can be passed through the locks. We will suppose six lockages per hour to be necessary, or N=6, we have n—4 (passing-places every fourth of a mile), t must, in advance of experience, be assumed. This time, t, is made up of the time lost in slacking the speed of the boat, in hauling it sidewise into the recess, in allowing the other boat to pass, in hauling out of the recess, and starting it forward to attain its ordinary speed, t can scarcely be assumed at less than five minutes (tV hour), and may equal ^ hour. We will suppose iV hour. Then, from the formula (^^ = N (v-j-nt), we find v=l; that is, the velocity of movement must he at least 3 miles per hour to permit six lockages in this time. If we double the number of passing-places, we find or the velocity of move- ment should be H miles per hour. Twelve turnouts per mile would require a rate of motion of 1 mile per hour. An increase in the value of t will necessitate an increase in velocity. If t=l hour, no possible velocity can enable 6 boats to pass out per hour, let the "number of turn- outs be w’hat it may. This substitution of ^ for t in the formula (N being equal to 6) will give v = 0; i. e., an infinite velocity will be necessary. This is as it should be, for the time of passing from one turnout to another, plus the time lost at turnout, must not exceed the time of one lockage. In this connection, it will be interesting to investigate tbe probable attainable rate of movement through a channel-way of this width and depth. 686 REPORT OF TilE CHIEF OF ENGINEERS. Some years since, M. Bazin made a series of experiments at the Pouilly tunnel of the Burgundy Canal, in France, which throws considerable light on this subject. This tunnel of Pouilly is 10.991 feet in length, with a water-way 2b..343 feet in width at the water-surface and 18,6928 feet at the bottom. The depth of water varied during the experiments from 7..546 to 7.71 feet. There was no current. The boats were 16.404.5 feet in width. “A dynamometer placed immediately behind the tow-boat received the tow-lines directly, and indicated each instant the force of traction. The velocity was determined by noting to seconds the time of ipassing bench-marks placed at distances of 100 meters. The results of these experiments presented numerous variations in their details, since the force of traction, more especially the velocity of movement, was incessantly modified through the influence of various causes — such as variable pressureof the steam in the boiler, the accidental movement of the boats in tow, which, not always follow- ing the axis of the tunnel, moved obliquely to it and even touched the sides ; the wave motion produced by the convoy, &c. “ The experiments were nine in number. The boats, when arranged in tows, were lashed to each other ?«s closely as possible — a j) re- caution usefql in diminishing the resistance.’’ The composition of the tows and the observation made for velocity, force of traction, &c., are given by M. Bazin in two tables, which are here consolidated into one and re-arranged. The French weights and measures are reduced to English units. Some of the items in the original tables are omitted as superfluous for our purpose. Table I. 4^ 1 o Composition of fleets. 1 1 1 velocity, in per second. k P > c c ii is 11 111 4 h: 11 "o c 1 11 111 is-l 3 1 Consists of one boat, rlraiip'ht .3'.9 A. 2. 365 2624. 7 1300. 4 .3076. 2 B. .3. 681 1968. 4 3769. 6 13875. 9 C. 2. 815 2624. 7 1807. 5 5088. 5 2 Consists of one beat, draught 4'. 4 D. 2. 139 2624. 7 1234. 5 2640. 6 E. 2. 345 3937. 0 1300. 4 3050. 2 3 Consisting of two boats, draught as follows : 4'. 2 and 3'. 46. r. 2. 526 3280. 9 1675. 4 4232. 0 H. 2. 692 2624. 7 2050. 1 5519. 0 I. 2. 638 2624. 7 1895. 8 5001. 2 4 Consisting’ of two boats, draught as follows: 4' and 4'. 4, K. 2. 302 2952. 8 1873. 8 4313. 4 and one empty boat. L. 2. 054 3280. 9 1366.7 2807, 3 5 Consisting of two loaded boats, draught as follows : 4'. 27, 4'.2 ; one boat lightly loaded, draught I'.llS, and one M. 2. 221 1312. 3 1454. 9 3231. 4 N. 2.319 1312. 3 1984. 0 4600. 9 empty boat. 0. 2. 920 4265. 3 2601. 2 7595. 6 6 Consisting of three rafts, draught as follows : 4'.59, 4'. 757, and 4'. 59. P. 2. 516 3280. 9 2777. 6 6988. 5 B. 2. 487 3280. 9 2909. 8 7236, 9 S. 2. 712 2624, 7 3350. 8 9087. 3 7 Consisting of three boats, draught as follows : 4'. 23, 3'. 67, T. 2. 100 3609. 0 3483. 0 9529. 6 and 4'.29. IT. 2. 099 2624, 7 2094. 2 4397. 9 8 Consisting of three boats, draught as follows : 4'. 49, 4M, and 4'.166. V. 2. 040 3280. 9 2006. 0 4091. 3 W. 2. 673 3280. 9 4078. 2 111.58. 1 X. 2. 349 3230. 9 2954. 0 6939. 9 9 Consisting of seven boats, draught as follows : 4'.!527, 4'.33, 4'.59, 4'.56, 4',0, 3'.969, 3h018, and two empty boats. Y. 2. 163 5905. 6 4166. 4 9032. 8 For the present we have to consider only the case of a single boat; that is, the first and second experiments. To facilitate our comparisons, the following table is compiled from the preceding information : APPENDIX V. 687 Table II. f bt+i ill 'C b/j® 'S'? S ^ § a?- .S S (D P. P -5+i O — gi p o o' 4) ot; ii s ^ o_ ® O P S . II c V- 5 ® A 'p « C b o'® I 3.9 !■ 4.4 64. 04 72. 12 1, 300. 4 3, 769. 6 1, 807. 5 1, 234. 5 1, 300. 4 20.3 53.8 28. 25 17. 1 18.0 1. 6125 2. 5097 1.9193 1. 4584 1. 5988 147.2 147.2 2. 295 2. 041 1. 190 1. 190 1. 9.35 1.714 On the central water-line the locks are proposed to be 24 feet in width, and the boats will then be about 23 feet 6 inches wide. If we suppose a draught of 6 feet 4 inches in a tunnel 34 feet in width, the water must be 10 feet deep to make the water section bear the same ratio to the submerged section of the boat as in the first ex^ieri- ment, and nearly 9 feet to make this ratio the same as in the second experiment. We will suppose these conditions to be the same, and that for the same velocity the neces- sary mean eftbrt of traction per unit of submerged section will be equal in the two cases. We can now compile the following table of velocities and resistances for a tunnel 34 feet wide, and the boats being 23 feet 6 inches in width, with a draught of 6 feet 4 inches. Table III. Letter of reference. Velocity, in niile.s, per lioiir, a.s per 'I'.ahle II. Necessary moan elfort of tractio]) i>er unit of submorged .sec- tion. 5 if cc ^ cl ?! •< Total effort necessary to pi’oduce given velocity, neglecting fractioiis. Batio of width of Avater-way to width of boat. Ratio of depth of water to draught of boat. Supposed dei)th of water. A 1. 6125 Founds. 20.3 143. 83 Founds. 3. 021 1.447 1. 579 Feet. 10 B 2. 5097 58.8 148. 83 8. 751 1. 447 1.579 10 C 1.9193 23. 25 148. 83 4. 204 1.447 1. 579 10 D 1. 4.584 IT. 1 148. 83 2. 545 1. 447 1.474 i 9 E 1. 5988 18 148. 83 2. 679 1.447 1.474 ! ^ This tunnel is somewhat wider in proportion to the width of the boats than the Pouilly tunnel, and of course has less proportional depth. ^ It will be observed that the last mean velocity of the second experiment (E) is but little less than the first mean velocity of the first experiment, (A.) and, though the sec- ond boat was drawing more water than the first, the mean effort per unit of section was considerably less than for the first boat. For some reason this second boat moved more easily than the first. We will take the second boat for the standard of comparison. If we suppose the resistances to vary as the squares of the velocities, one mile perhour would require 8.022 pounds traction per unit of section according to D, and 7.03 pounds according to E. Seven pounds per square foot will be a total of 1,041.8 pounds force of traction, the necessary amount to move a boat 1 mile per hour. ' It was proposed in this single tunnel, with passing-places, to have a timber tow- path 5 feet 9 inches wide, in case horse-power were used on the canal ; this tow-path to be omitted if steam-power were adopted on the canal instead of animal-power. The Ordnance Manual (page 472) gives 120 pounds as the mean effort exerted by a horse drawing a cart or boat walking, and working 8 hours per day. McAlpiue men- tions that, according to experiments made in France, a horse can exert 143^ pounds for 6 days. These figures refer to a more rapid motion than 1 mile per hour. Trautwine 688 REPORT OF THE CHIEF OF ENGINEERS. gives 250 pouucls as the power of traction which a horse can e±ert traveling 1 mile per hour. But we have here a timber tow-path. The footing cannot be as good as upon earth, even by placing some of this material on the path, (which would cause it to rot out quickly,) and several horses would be necessary, arranged in single file. These disad- vantages would seriously decrease the effective power of the horses. Our boats have 148.83 square feet of su])merged section, loaded to 6 feet 4 inches. One pound per square foot seems to me to be all the effective force we could expect under the circumstances from each horse. Assuming this to be so, at least 7 horses would be required for each boat to move 1 mile per hour. At the turn-outs these 7 liorses, with drivers mounted, would have to pass the horses of the boat in the turn- out. This manoeuver on a narrow tow-path must be tedious and difficult. The time allowed at turn-outs (fg hour) is surely as short as with safety can be assumed. One mile per hour, we have seen, will render 12 turn-outs per mile necessary. Each being 140 feet long, their aggregate length per mile would be 1,680 feet. Thus nearly one-third the entire length of the tunnel would be excavated to a width sufficient for passing boats everywhere, which, with the expense of finishing off the ends of the recesses and the expense contingent on changing from one section to another in the excavation, would probably make fully one-third the difference of cost between a sin- gle tunnel 34 feet wide and the tunnel of 52 feet width, as originally proposed. The uncertainty of action of the 34-foot tunnel would still remain to be considered, together with its disadvantage of not admitting an increase in the trade of the line in excess of 6 boats per hour, (i. e., 3 each way,) A current through the tunnel from west to east would somewhat change the condi- tions. The eastward boat would have the benefit of this current, and would reach the turn-outs in time probably to pass into the recess before the arrival of the west- ward boats. The westward boats would have no delay at turn-outs except in the diffi- culty of passing horses. Supposing no delay, (as might be the case if a recess were provided for horses beyond the recess for the boat.) the formula wmuld reduce to = N r, and with a rate of movement of 1 mile per hour, only six turn-outs would be necessary per mile to provide for six lockages per hour. I estimate that a velocity of more than one-half mile per hour will be necessary through a channel 34 by 9 to supply water for the canal eastward to Covington. The westward boats must stem this cur- rent, and to make 1 mile actual progress would require about the effort of traction corresponding to the second part of the second experiment. Assuming 148.83 pounds to be the effective effort of each horse, 18 horses would be necessary ; an unmanage- able number in such circumstances. Fewer horses will give less sj)eed and render more turn-outs necessary. It may be said that the westward boats will be more lightly loaded than the east- ward ones, (i. e., to less than 6 feet 4 inches,) and therefore will offer less resistance than that above mentioned. Generally this will be so. But a large transfer of iron- ores from east of the Alleghanies to the west is to be expected. A single heavily- loaded or slowly-moving westward boat will modify the time of passage of every eastward boat which it meets while in the tunnel, being one for each turn-out. If we suppose but four turn-outs to the mile, and that heavily-loaded boats shall succeed each other at intervals not greater than ten hours, the whole traffic in the tunnel must conform in time of passage to these heavily-loaded boats. A current of one- half mile per hour must render the checking of the speed of the eastward boat and the hauling of it into the recess a difficult manoeuver in a contracted space such as this 34-foot tunnel. It cannot be done quickly. About 300 pounds traction would be necessary to hold the boat in this channel-way against a current of this velocity ; more than the effort which two horses put forth in towing at a walk. Without expanding the sub- j ect further — already perhaps too long drawn out — I would state my opinion as decid- edly opposed to the tunnel with passing-places for single boats, the boats being moved by animal-power. It should be remarked, however, the original proposition contem- plated the use of boats which would offer less resistances to movement, as the locks were assumed but 20 feet in width. If steam-power be used on the canal, the tow-path is dispensed with, and circum- stances change. The eastward current will aid the eastward boats, and they should, without doubt, enter the turn-outs, and probably would be able to complete this ma- noeuver at least in time to enable the westward boat to proceed without delay. The formula = N v will express the relation existing between the velocity of the west- ward boats, the number of turn-outs per mile, and the number of boats which can pass out of the tunnel per hour. With 4 turn-outs per mile, the net velocity of the west- ward boat must be 1^ miles per hour to provide for 6 lockages in that time. The ac- tual velocity against the half-mile current must be 2 miles per hour. With this ve- locity, the resistance per square foot of submerged section would be 28.1 pounds, according to the first part of the second experiment, (D,) and 26.7 pounds, according to the second part (E) of the same experiment. The first experiment gives 28J pounds for a velocity of 1.92 miles per hour, water-way being 10 feet deep. We will suppose APPENDIX V. 689 28 pouuds to be correct. The total resistance for a boat clrawinj:^ 6 feet 4 inches will then be 4,167.24 pounds. Two miles |)er hour is 176 feet per minute. The work of moving the boat at this rate will be 734,434 foot-pounds per minute, or 22^ horse-power (effective force) must be employed. Lagrene says that the co-efficient of the useful effect of the screw varies between 0.42 and 0.64, taking tlie work of the pistons of the engines as unity. — (Experiences Dynamometriques, par M. Taurines, 1859.) This is for marine-engines. Labrousse states (Traits de Touage sur Chaiue noyee) “that in our canal, (France,) where the dimensions of the lock-chambers compel the use of boats with certain particular con- ditions of form, any system of ^ * screws * * * -will not realize more than 25 per cent, of the motive force.” The loss in this tunnel must be still greater. It is therefore more than probable that the velocity necessary, with four turn-outs per mile, cannot be obtained with engines at all economical for the navigation of the open canal. It may be, though exceedingly doubtful, that the power used on the canal could gen- erate a lower velocity, which, with a permissible number of turn-outs, would provide for six lockages per hour. A number of turn-outs permissible in view of expense is here meant. Without discussing the question, I would state an objection to any sys- tem of tJirn-outs, which becomes more grave as their number is increased. If the tun- nel were worked systematically and to its full capacity, each compartment, as noticed above, would contain a boat, adjacent compartments being occupied by boats moving in opposite directions. Each boat, when arrived at a vacant turn-out, should with- draw into it and remain till passed by a boat moving the other way. It could then proceed. This alternation is positively necessary, and, as long as the working be. full and regular, would operate very well, but if there be a failure in either of these condi- tions, confusion must necessarily ensue. The tunnel proposed would have thirty or more turn-outs, supposing four per mile. Each of the numerous delays incident to any traffic will have its effect in increasing the confusion within the tunnel. One boat moving irregularly will communicate this irregularity, in a greater or less degree, to sixty boats. In a time of shck trade boats would not present themselves at proper intervals at the portals. Some compartments would then be unoccupied by boats, unless indeed all the boats withjn the tunnel be compelled to await the arrival of those behind them. Ourside certain limits this could not be thought of. The captain of each bo it, when arri vmd at a turn-out, must know whether to withdraw into the recess or to pr »ceed. A mistake would be very difficult of correction. We cannot hope that thirty turn-outs will answer our purpose, and a greater number will increase tlie diffi- culties. This objection should of itself prevent the adoption of passing-places for single boats in a tunnel of considerable extent. A single tunnel wide enough for passing everywhere, or two tunnels of single width, would remove this objection. Though the movement would be more slow than on the open canal, the boats could pass out one end as rapidly as they could possibly enter at the other. The importance of decreasing the first cost impels us to provide for an or- dinary trade at least, without the construction of the second tunnel, or a large tunnel, if it can be avoided. The movement of boats in fleets seems to offer the only chance for a solution of the problem. In the formula a = N (v + ni), we can suppose N to be a number of such fleets, and t, instead of the time lost by a single boat at each turn-out, will be the time lost by the entire fleet, and, if the eastward current be sufficiently rapid to enable an eastward fleet to arrive at and withdraw into a turn-out before the arrival of the westward- bound boats, i becomes zero, and the formula reduces to w = Nr for the westward fleets. Preparatory to the solution of the problem, the experiences at the tunnels of the St. Quentin Canal, in Belgium, and at the Pouilly tunnel, in France, are of great value. M. Lemoyer contributed a memoir to the Auuales des Pouts et Chaussdes. 1863, on the towage of boats in the tunnels of the St. Quentin Canal, containing much valuable infor- mation. It is here condensed, those portions being omitted whicn do not immediately concern us in our investigations. “ There are two tuimels (Riqueval and Tronquoy) on this canal, separated by a short distance. The Riqueval tunnel is 18,603 feet in length, and the other is 3,606 feet. The water-way was at first 17.06 feet in width, and 7.546 feet in deqith, with a solid banquette on each side 4.6 feet wide, and raised about 2ifeBt above the ordinary water- level. From the opening of this line, in 1810, to 1857, boats were towed through these tunnels by men and women, eight or ten being assigned to each boat. The boats for- merly drew nearly 5 feet of water, and thoiigh all the locks bad a miiiimutti width of 17.06 feet, the greater part of the Flemish boats had but a width of 14.44 feet. The privilege of the tunnel was free to all. The slowness of movement was such that the boats generally required twenty hours to traverse the Riqueval tunnel from portal to portal. Transit in the same direction could occur only on each second day. “ The increase of commerce necessitated the adoption of some method of traction which would reduce the delay incident to hauling; and January 20, 1840, ten men were employed for each twenty tons, in order to increase the rate of movement and to 44 E 690 REPORT OF THE CHIEF OF ENGINEERS. double tlie utility^ of the canal permitting: transit each way each day. Competition with railroads made the enlargement of the Flemish locks necessary to admit the wider boats, and the draught of the boats was increased to 5.91 feet, and the depth of the water throughout the line to 6.56 feet. With this increase in the draught of the boats, the difficulties of traction became so great that hauling by men was practically impos- sible. The time of transit, which before this modification of the boats was seven or eight hours, soon incroased to sixteen and eighteen hours. By degrees the men, from exhaustive labor, refused to continue work, and an increase of wages could not prevail on them to remain in a service which was beyond their strength. Foreseeing an immediate abandonment of this means of traction, the management tried various means to assure the towing of boats. A trial of steam-towing had demonstrated that its employment was here inapplicable, with the defective smoke-consumers then in use, on account of the inconveniences arising from the gases generated by combustion and because of the ill-effects of frost within the tunnel, which is excavated through chalk, and is but partially lined with masonry.* “ When the boats were towed by hand, doors were used at the ends of the tunnel to diminish the action of the frost, and the shafts had been hermetically closed. Deprived of this mode of traction, the management had recourse to horse-power. The banquettes were not provided with guard-rails, and the use of animals on such a narrow tow- path was not thought of without providing by some means to prevent accidents. Mov- able guard-rails were then adopted, which removed all danger. Two large transverse bars were securely lashed with ropes near the bow, by means of which the guard-rails were attached to the boat, and moved with it. A horse on each banquette was tied to the fore part of the guard-rail and the tow-line was attached to the boat. “A fleet of fifteen boats was first tried, towed by 30 horses. The time of transit ex- ceeded 14 hours. Two other trials not more successful. ‘‘ Finally, fearing that this low rate of speed was due to the fact that the horses were unaccustomed to this kind of labor, twenty teams perfectly broken by daily use in the approach cuts were used, and placed in charge of the best drivers. “ The number of superintendents was doubled, yet, notwithstanding all precautions taken, the experiment failed completeljq as the convoy required 13 hours to traverse the tunnel. “It was found that the low rate of speed, by compelling the horses to take short steps, neutralized most of their power, and the result produced was altogether out of proportion to the fatigue endured. More powerful means were necessary. “ Out of a number of methods proposed, but one appeared worthy of trial. It con- sisted simply of a boat fitted with a platform, raised to clear the banquette, on which horses traveled to turn a capstan. A rope or cable attached to some point ahead, and making two turns round the shaft of the capstan, provided the means of forward move- ment when the capstan was turned. This device was of great value, as the increase of power permitted the formation of tows of 30 to 40 boats. The time of transit was even now from 10 to 12 hours. “ Improvement in the details produced no acceleration of movement, since the resist- ance due to the contracted width of the tunnel could not be diminished, and it in- creased in great proportion to the increase of speed. “The water-way was next enlarged by the removal of one of the banquettes, giving an increase of 4.6 feet in width. As a further improvement the ropes w^ere dispensed with, and a submerged chain was put in use instead. Eight horses were used to tow large fleets, but this number might be diminished according to circumstances. The chain after three years’ use has suffered no deterioration, and none of its links have been broken. The arrangement enables each horse to tow about one thousand gross tons, moving regularly with a velocity of about f of a mile per hour. This mode of traction has reduced the cost so materially that ithe toll is less per ton per mile than on any other portion of the canal, or even on the river navigation connected there- with. The time of transit was reduced to about seven hours for ascending fleets, and to little more than five hours for those descending.” In 1868 1 steam-towage with submerged chain, theretofore considered impossible, was inaugurated. M. Bazin has given in a note (Annales des Fonts et Chauss^es, 1868) very interest- ing details on the use of steam-towage established at the Pouilly tunnel of the Bur- gundy Canal, in which note were published the experiments mentioned in a previous part of this report. From this note the following information is extracted : “ The tunnel of the Burgundy Canal, in which a system of steam-towage has just been established, has not the exceptional length of that of the St. Quentin Canal. The trade of this line, paralyzed even to the present time by the condition of the river * Steam-towing has since been successfully inaugurated. — T. 'f. t Letter to Prof. G. L. Andrews, U. S. M. A., from Mr. E. Malezieux, Chief Engineer of the Corps des Fonts et Chauss^es, transmitted to General J. G. Barnard, Corps of Engi- neers, U. S. A. APPENDIX V. 691 Yoiine, the improvement of which is not yet completed, is still less active; and the tunnel is on the least frequented portion, where the annual carrying trade is only about 120,000 tons. Navigation on this account was none the less subject to serious embar- rassments which a more active trade would immediately aggravate. The tunnel is 10,991 feet in length, with an approach cut 2,953 feet long at each end. The approach cuts, like the tunnel, are but wide enough for one boat, being 21.98 feet at the water- surface, and 20.38 feet at the bottom. “The water-way of the tunnel is 20.38 feet wide at the water-level, and 18.7 feet at the bottom. “The water in the summit-level is constantly 7.2 feet to 7.87 feet in depth. The length of the portion with single widths, comprising the tunnel and approach cuts, is 3.2 miles. “The rule in operation before the establishment of a system of towage allowed the boats 6 hours to pass over this distance ; the convoys approaching from the side of the Yonne entered at noon and midnight, and those coming from the water-shed of the Sa6ne, at 6 o’clock, morning and evening. The motive power was furnished by men to the number of 4 to 6 to each boat. The tunnel having no tow-path, these men hauled on a chain attached along one of the sides. “The establishment of a more powerful mode of traction became urgeut, especially in view of the probable increase of traffic when the canalization of the Yonne should be completed. A decree of April 28, 1866, authorized the creation, at the expense of the state, of a system of steam-towage, which was placed in operation on February 5^ 1867. “The tow-boat employed at the Pouilly tunnel was constructed according to the sys- tem of M. Bouqui4. This system, of which the small one in use at Paris, in the last level of the St. Martin Canal, is a specimen, is distinguished from other methods in this, that the chain, instead of passing over the middle of the boat, rests simply on a pulley at the side, passing over only a portion of its circumference. The groove of this pulley is provided with recesses in which the links of the chain are engaged, which re- quires that the links should be of a uniform size. The lateral position of the pulley has^ for very narrow tow-boats, as those operating in a tunnel of single width should be, the advantage of clearing the deck, which the passage of the chain over the middle of the boat would incumber to an extent frequently dangerous and embarrassing while maneuvering.* “The engine is high-pressure and condensing,! of fifteen horse-power. Motion is transmitted to the shaft of the pulley by means of a belt and gearing under the deck, which thus remains perfectly free. The gearing permits twm different velocities of the pulley, according as the boats in tow be heavily or lightly loaded. These velocities are as 3 to 5. The total length of the boat is 72.2 feet, and its width 10 feet 8 inches. It cost $8,074.50. “The steam tow-boat at the Pouilly tunnel is probably the first employed for this purpose in France, and the experience here obtained furnishes useful data on the special difficulties of its use in like cases. The resistance to traction, on which a connected series of experiments has not been made, is very great, and the movement of convoys creates very complicated undulatory motions throughout the whole extent of the level. This great resistance requires that a large margin should be allowed for the estimated powder to be given to the tow-boat. The w^ater displaced by the progress of the boat not having, as in a w^ater-way of great width, room to flow off freely at the sides, is obliged to escape with great velocity through the narrow space between the boat and the sides and bottom of the cut, thus causing a permanent difference of level from stem to stern of the boat, which differ- ence of level is greater as the dimensions of the boat approach those of the canal. In experiments made by us this difference at times was nearly 8 inches. “ The progress of a convoy of boats in a tunnel of great length causes movements in the entire mass of the level. As soon as the convoy enters the tunnel, it presses the ■water before it, which pressure is transmitted forward in the form of a wave, which traverses the length of the level to its extreme end, from which it is reflected, and re- turns toward the convoy, reaching which, it passes on to the other extremity of the level, from which it is again reflected, and so on with gradually-diminished height. It is besides clear that any variation in the velocity will create secondary waves similar to this. An increase of the velocity will increase the difference of level, producing a new wave, the motion of which will be in the same direction as that of the convoy. “ On the contrary, a slowing-up of the boats will allow the water to flow backward to attain its equilibrium by passing under the boat, causing a new wave, which will be *' This method permits the chain to be readily taken up or thrown off', which cannot be done if the chain passes over the middle of the boat. — T. T. tThe condenser is absolutely necessary in a tunnel, as, because of the slowness of movement, the escape of the steam immediately produces complete darkness.— M. Bazin. €92 REPORT OF UHE CHIEF OF ENGINEERS. propagateil in a direction opposite to the motion of the convoy. Variations in the sec- tion of the canal will also produce effects of the same kind. These causes collectively give rise to very complicated phenomena. We had no very certain data to determine the power necessary for a tow-boat. Some trials made in hauling one, two, and three boats by men in the tunnel and horses in the approach-cuts would seem to give for the value of the effort IF) of traction, necessary to produce a velocity (r) with (w) loaded boats drawing 4'. 6 feet. F = 159 n which, expressed in horse-power, is P = .2890 n v^*. According to this formula, a power of fifteen horses should suffice to tow a convoy of seven boars with a velocity of nearly 2 feet per second. It was even very probable that this result would be exceeded in practice, since it could be affirmed a jmori that the force of t action should increase in a loss ratio than the number of boats in tow.’’ These estimates liav'-e been confirmed by careful experiments, the results of which have been giv'en in Table 1. M. Bazin, in discussing the results of these experiments “ without an absolute rule, which tile nature of things does not admit of,” groups the results into a very sim- l)le, approximate formula, by remarking that the “co-efficient of traction [ — ; ) of the first boat is about double that of each succeeding one in the same fleet.” This formula oisoning, of which the first symptoms are vertigo, accompanied wirh violent headache, weakness in the legs, and sometimes vomiting. The crew of the tow-boat have several times, after traversing the tunnel, felt these effects. However, by carefully watching the fire, taking special pains not to overload the grate when passing through the tunnel, these troubles may be almost avoided. “ The sulphurous acid, notwithstanding its irritating action on the organs of respira- tion, x^resents much less danger, but it is imx30ssible to prevent its formation, all cokes containing some traces of sulphur. “There is also produced a small quantity of sulphureted hydrogen, the presence of which is explained by the passage through the incandescent coke of the small jet of steam used to create a draught. This gas is in part decomposed by the moist air of the tunnel liberating the sulphur, which remains in a state of suspension in the smoke, giving to it a very peculiar milky appearance. When the gas accumulates on the tow-boat, the decomposition of the sulphureted hydrogen even suffices to cover certain objects with a light coat of sulx)hur powder. The cause of the accidental accumulation of gas being known, by what means can a remedy be apx^lied? “ The only means insuring efficacy would be evidently to establish at the two ex- tremities of the tunnel ventilating furnaces, with movable partitions permitting the current to be reversed at will, so that it might always be ox^posed to the motion of the boat. This radical solution, admissible in the case of a very important traffic, would have necessitated, unfortunately, at -the Pouilly tunnel, sacrifices out of proportion to the very restricted trade, and it became, therefore, necessary to have recourse to means less certain without doubt, but more economical in application. The curbs of the shafts have been raised 9 feet 10 inches above the surface of the ground, and the heavy cast-iron gratings have been removed, which obstructed about half their open- ing. This modification has sufficed to notably increase the circulation of air in the tunnel. The shafts are 4 feet 11 inches in diameter ; their axes are 8 feet from that of the tunnel, so that their openings do not debouche from the crown of the arch, where the gases have a tendency to accumulate. It would then be advisable, in order to ren- der the draught more energetic, to so widen the lower openings of the shafts as to join them directly with the crown of the tunnel-arch. This oxieration performed at one of them seems to have produced a good result. Another very simple precaution has re- stored the confidence of the crews, who were much frightened by these accidents, tlie cause of which escaped them. “ The accumulation of gas being due to an accidental coincidence of motion between the tow-boat and the current of air in the tunnel, it was to be presumed that by ar- resting the motion of the convoy for a few moments and closing the smoke-stack the current of air would resume its action and dissipate the gas. A trial confirmed this supposition. Instead of accelerating the movement, as was at first injudiciously done by the crew of the tow-boat, a few minutes’ detention sufficed for the current of air to carry the smoke and gas far in advance, and it rarely happened that the accumulation again became very troublesome in the course of the same passage. “An im^jortaut amelioration, it would seem, might be made in the machine itself; the addition of an apparatus by which the gas issuing from the smoke-stack could be cooled and even purified in part. This might, without doubt, be accomplished by clos- ing the chimney and forcing the gas by means of a bellows into a reservoir, where it would have to traverse a circuit of a greater or less extent in contact with tlie water of the canal, and then expelled at the surface-level. At present the warm gases tend to accumulate at the crown of the arch, there forming a local atmosx^here about the heads of the men employed on the deck of the tow-boat. expelling them in a cold condition at the water-surface, these gases, for the greater part more dense than the air, would have little tendency to rise, and would become intermingled with the surround- ing atmosphere by the movement of the boats. Their contact with the water in the cooler would have the additional effect of dissolving a great part of the sulphurous acid. “This apparatus, the study of which has been asked of the constructor, not having as yet been established, the amelioration which we indicate has not received the sane- 694 REPORT OF THE CHIEF OF ENGINEERS. tioD of experience. However, and althoiif^b the problem of ventilation has been but imperfectly solved at the Ponilly tunnel, the service has j^one on regularly for a year back ; and even if the men employed in this service are sometimes a littH incommoded be fears which they had conceived in the beginning have been completely dispelled.” M. Bazin, writing January 25, 1875,* * says : “ The service is performed with a perfect regularity, and it is exceedingly rare that any one is inconvenienced by the gas. This result is due to the experience acquired by the personnel and especially by the vigi- lance of the engineer whom we have had since September, 186S. The personal influ- ence of a good engine-man is capital. The one we had in the beginning allowed mat- ters to take their chances, and it is without doubt to his negligence we must attribute in great part the sad accident which happened in the summer of 1868. This accident, account of which was rendered to the administration, and probablj^ with which you are therefore familiar, took place under the following circumstances: The crew of the tow-boat and even the boatmen found themselves very much incommoded by the gas, which a boat with an elevated cargo contributed to retain on the convoy. The men collected on the tow-boat, detached the tow-lines, and abandoned the boats in the tun- nel. Unfortunately they had not been able to bring with them a boatman who was asleep in his cabin. This man, subjected in his sleep to the deleterious action of the carbonic oxide, against which he had not been able to bear up, had lost consciousness, and the others could not succeed in carrying him to the tow-boat. When they returned to the abandoned convoy no further efiect of the gas was felt, but the unfortunate boatman, in spite of all efforts, died on the following day without return to conscious- ness. “The new engine-man, who has been employed since this accident, is a very careful person, who has himself made successive improvements in the details of the grates, which in the first tow-boat were certainly defective. The intervals between the grate- bars, the dimensions of the openings, have been so modified as always to have a clear fire. “The first boat sent by M. Claparede had a steam-blast throwing a jet of steam be- neath the grate. Perceiving a j) reduction of hydro-sulphuric acid, I caused the jet to be placed in the smoke-stack, but the crew made no further use of these blasts, which they accused of having contributed to the previous accidents. “According to the exj)erience of the engine-man, it is always necessary to enter the tunnel with a clear fire. If they enter with a fire imperfectly lighted and with too much coal on the grate, the action of the blast employed to hasten combustion wdll be dangerous. It is probable that a too rapid current of air through the flue forces out the carbonic oxide before it has had time to burn. Finally, although the service operates perfectly at present, the question of ventilation has not much advanced. The tonnage which we have to move is too small to render costly experiments possible. I know not whether this question has been more fully studied on the St. Quentin Canal, where the importance of the traffic would perfectly justify the construction of special apparatus. “ Our shafts have little depth. The neck of Pouilly is very depressed, and the maxi- mum depth is 131^ feet, which is very little for a tunnel so long. The intervals vary from 524 to 656 feet. In the beginning these shafts were located in couples — that is to say, there were in each 656 feet two shafts 131 feet apart. A certain number of them were closed after the completion of the works, and generally there remains but one shaft of each couple. #***#*# “It is to be regretted that, operating on such a small scale, we cannot undertake a complete study of this question, difficult and now obscure.” In the letter relating to the St. Quentin tunnel, previously referred to, is the follow- ing valuable information : “ In 1868 steam-towage with submerged chain, heretofore considered impossible, was inaugurated. Since then it has been saccessfully organized aud perfected. It is at the present time in operation with regularity and success, by means of two tow-boats, each of which traverses twice per day half the summit-level, being about l:;f miles going and returning. When a third boat, now in construction, will be finished, the regular- ity of the service will be perfect, as a tow-boat in need of repairs can be temporarily replaced. “The long tunnel of the St. Quentin Canal, which extends from Macquincourt to Riqueval, presents a development of 18,603 feet. Shafts located at distances of 100 meters, and numbered from the Riqueval end, were excavated in its construction. The sides of the tunnel are protected by a masonry revetmeut for 9,515 feet ouljq being little more than half its length. Throughout the remainder the chalk forms the sides of the gallery. The limestone there is very much affected by the frost, and it has been con- sidered indispensable to stop up all the shafts, and to complete the closure by a system * Letter to M. Malezieux, transmitted to-Brofessor Andrews, U. S. M. A., for General J. G. Barnard. APPENDIX V. 695 of turniuor-floors, by means of which the tunnel may be closed at the extremities dur- ing freezing weather. The ventilation had been imperfect and the employment of steam in towing of boats had been despaired of. It has been accomplished, however, by the following precautions: “ 1st. The 9 shafts, Nos. 5, 14, 18, 25, 29, 32, 37, 44, and 49, have been reopened. “ 2d. Their upper extremities have been extended by chimneys 3 to 4 meters in height. “3d. The tow-boat for traversing the tunnel burns coke, and the use of a powerful blast renders the production of smoke almost nothing. “ The shafts are of variable depths, from 130 to 330 feet. Thanks to these precau- tions, the passage of the tunnel is effected without any person ever complaining of having been inconvenienced. As to the action of frost on the non-revetted sides of the tunnel, its effect is scarcely sensible. When the temperature lowers and the traffic is arrested by ice the doors of the tunnel are closed, and the circulation of air almost ceases.’’ The experience of these two long canal-tunnels demonstrates a number of important facts. Among the number are : 1st. lliat steam-towing has been successfal. 2d. That to overcome the difficulties of veutilatiou, special devices in the machinery of the tow-boat have been necessary. 3d. That coke should be used. 4th. That it is desirable to have as few steamers in operation as possible at one time within the tunnel. Special steamers will therefore be necessary, and the boats should move in fleets. This is true, eveu if the tunnel be wide enough for passing everywhere, or two single tunue's be used. lu the latter case, if the boats were to use their own steam-power, there would be with a full trade probably more than thirty boats in each tunnel, all steaming at once, while, if the boats were moved in fleets, there would be but one steamer to each fleet. Besides, these tow-boats would have special appliances to decrease the quantity of smoke and gases, which would be too much to ask of all boats doing business on the canal. ^ Believing the necessities for special tow-boats to be evident from what precedes, further discussion is not attempted. To move boats in fleets, the method by submerged chain is undoubtedly the best. There is comparatively little loss of power. The method is very simple in its arrange- ment. It consists of a chain or cable, (wire,) laid on the bottom, securely attached at the ends, with a tow-boat arrange I to maneuver along it. What is most important, it has been put to the test of actual experience, and is there- fore not at all experimental. In France this method has received extensive applica- tion on the rivers, and is also in use in a portion of the Erie Canal, under the name of the Belgian system of towing. My opinion as at present formed is in favor of using a chain, on account of the facility of repair. A break in a wire cable would require an expert to splice it. This would be difficult to do if there were not considerable slack available, as there would not be in this instance. The chaius used are made in sectionSj connected by links, which may be readily opened. The section containing a broken link may then be removed at any time, and a new section put in its place. A supply is always kept on the tow-boat for emergencies. For greater security I would recommend that the tow-boat be provided with a pul- ley on each side, and that two chaius be laid on the bottom at a distance apart equal to the width of the tow-boat. This would be very necessary in a tunnel with a cur- rent, so that if one chain broke the fleets would still have one to hold to; otherwise all the boats eastward of the break would be washed into the basin of the eastern approach. The chain has the further advantage of greater friction on the bottom, as compared with the wire cable. The chain will have an important use in assisting, by the application of proper brakes, the retardation of the descending fleets preparatory to hauling them into the recesses. Two chaius will be advantageous in this maneuver. The chain always has a tendency to draw to the inner side of a bend, and if it did do so would prevent the progress of the tow-boat. A rapid current corrects this tendency, or, if the water- course be wide, the steering power of the boat will compel the chain to remain in the channel. Within a tunnel we can depend on neither of these aids, and there must be no tendency of this kind. PEUFECT STRAIGHTNESS IS NECESSARY. Attention is now invited to sheet No. 1 of the summit division. Mr. Hutton’s pro- posed tunnel of 1870 is laid out on a curve, so as to decrease the depth of shafting, by placing them farther toward the valley than the surveyed or experimental line. By increasing the depth of shafting, this location could be made perfectly straight. No straight line can be located within the ’units of the survey of 1874 from Brush Creek 696 EEPOET OF THE CHIEF OF ENGINEEES. to tlie Greenbrier River direct without great depth of shafting or great distances- between them. From the initial point of this direct line (designated Northern line, 1874) to the south fork of Howard’s Creek the ravines fall to the north of the line, and a very good location can be obtained on that side, straight to the Greenbrier Mount- ain. The best line for tunneling this mountain is by way of the two ravines imme- diately south of the line. But this line makes an angle with any straight line from Brush Creek to the Greenbrier, and is therefore not available for our purposes. If we limit the depth of shafting to about 600 feet, the best straight line through here is from Brush Creek to a point in that ravine debouching into the valley of Howard’s Creek, north of the line, and heading near station 507 + 33. We would have on this location one shaft (eastern) GOO feet in depth, and one (western) 520 feet, with a distance between of 9,450 feet. Estimating 30 feet per month for shaft-excavation, and 100 feet for tunnel-heading, the time of completion would slightly exceed sixty-six months. The length of the tunnel would be about 50,500 feet. On the southern line, from Brush Creek to the south fork of Howard’s, the ravines- fall to the south. Beyond the railroad, there is a ravine south of the line, heading near station 368 -f- 50. On the other side of the crest the ravines fall to the north, but shafting may be obtained south of the line, and lower down the ravines than for Hut- ton’s location. A straight line may then be obtained from Brush Creek to Howard’s Creek, south of the surveyed line. This line is chosen for the following reasons : It has much more advantageous shafting than either of the others as to depth, distance be- tween shafts, and proximity of location to railroad ; it requires about 9,000 feet less tunnel than the line from Brush Creek to the Greenbrier direct, and while the tunnel is somewhat longer than Hutton’s proposed tunnel of 1870*^, there is saved more than 13,000 feet of canal on the eastern slope. The necessity of towing the boats through the tunnel in fleets being admitted, there must be a basin of suitable dimensions at each end. As the line must enter the valley of Howard’s Creek below the level of the stream, on account of water-supply, the basin at this end is readily obtained by biyldiog a dam at a proper point, and a basin may be obtained in Brush Creek Valley in the same way. The elevation of the tunnel, which is the summit, should be no-^ fixed. The eleva- tion of 1,700 feet above tide was recommended by Mr. Lorraine, and adopted for Hut- ton’s location of 1870. On account of the expensive approach in Howard’s Creek, it has been suggested to raise the summit 20 feet. (See Mr. Latrobe’s letter to Chief of Engineers, supplementary to the report of the board of engineers, March 19, 1874.) Various considerations enter the discussion of this question. If the lower level be ^ adopted, the basin in Brush Creek Valley (see sheet No. 2, summit division) may be" formed by a dam a short distance below station 77 -4- 50, and the tunnel should de- bouch near the u])per end of the pool. This pool will have about 1,800 feet effective length; its total length will be somew^hat more. If the higher level be chosen, the dam may be located at station 89, giving a pool of abouf, the same effective length as the other. There is scarcely enough difference in the capacity of these pools to have weight in the decision. A basin of necessary capacity for either level may be obtained in Howard’s Creek. The valley of this creek (see sheet No. 3, summit division) below Hart’s Run to Caldwell’s station is narrow and tortuous. Near the point where Hutton’s line of 1870 debouches is a rock-spur 80 to 100 feet above the vuilley, and causing it to make an abrupt bend. From station 582 -f- 47.8 up the valley, for about 1.000 feet, high ground, having the character of a bluff, comes very close to the line. The level of the creek near here is about 1,744 feet — too high to permit it to enter the canal at the higher level. For either level, the cutting will be deep and long. It must also be wide if it pass around the bend ; in which case the creek should be passed through the spur. To avoid this, I recommend that the creek bo passed over the cut, and that the spur be tunneled for the canal. The canal and creek wall then connect below. If we suppose the creek-bottom 4 feet above the crown of the aquedu3t-arch, and that 20 feet be as little as can be admitted between the water-surface and the crown, then the w'ater-surface can be no higher than (1720.) This is, therefore, taken as the superior limit of the elevation of the summit [?. e. with “canal bottom ” at (1713.)] We can now locate the funnel-lines for these two levels, (1713) and (1700.) It is de- sirable to have the approach cut in the valley of Howard’s Creek straight, and in the prolongation of the tnunel, so that the cut may be no wider than the tunnel and the towage by submerged chain may be used from one basin to the other. The low bluff above station 582 -f- 48 should be avoided, and in Brush Creek Valley the tunnel should debouch so as to utilize the basin corresponding to the level taken. The eastern portal for the (1700) level is taken near the upper extremity of the 1700 curve. The entire pool will then be below. Line “A” is the line recommended for this level. For the (1713) level line “B” is recommended. This line has almost the identical * Hutton’s tunnel is 40,380 feet in length. 'Ihe tunnel now recommended is 41,505. APPENDIX V. 697 lo-ation of approach in Howard’s Creek as the (1700) level, (see sheet No. 3, summit di- vision) bnt debouclies in the valley of Brash Creek at the upper extremity of the (1720) curve. These lines, as shown on sheet No. 1, are the lines of shafting. If the lower level be taken, it is intended the tunnel and approaches shall be north of the shafts, while the opposite is intended for the higher level. The basin in Howard’s Creek Valley is formed in each case by a dam jurt below sta- tion 106 -f 73 of the Howard’s Creek line. Each of these lines has advantages and disadvantages peculiar to itself, and it is thought best to place a statement of them side by side the better to judge of the merits of each. Upper level. Lower level. Length of main tunnel, 41,505 feet. It is probable that shaft No. 6 of this line is unavailable from depth and diffi- culty of approach. By shafts Vis, and 7, the excavation will require 57^ months. I5y shafts Nos. 5 Vis, and 8, the excava- tion will require 62.13 months. By shafts 1 and 3 Us, Lewis Mountain will be tunneled in 47.5 months. By shafts 1 and 3 Us, Lewis Mountain will be tunneled in 48f months. Pumping will stop at shaft No. 1 in eleven months, and the eastern heading of Lewis Mountain will be open to the portal. Length of main tunnel, 4.3,040 feet; be- ing 1,540 feet longer than for higher level. By using shafts Nos. 2, 6, and 7 of this line, the tunnel may be excavated in 50| months. By shafts 5 Us, and 7, the excavation will require 59f months. By shafts Nos. 5 Us, and 8, the excava- tion will require 64| months. By shafts 1 and 3, Lewis Mountain wilt be tunneled in 51| months. The eastern heading of Lewis Mountain will not be open to the portal till the end of the twenty-second month. This is quit© an advantage to the upper level. It is thus seen that, for time of excavation, the only apparent advantage possessed by the lower level, exclusive of lockage, is the availability of shaft No. 6, and less time for the excavation of the headings between shafts Nos. 3 and 4, and between shafts Nos. 4 and 5, or 5 Us. In the estimates made for times of excavation, 30 feet per month is allowed for shaft-excavation, and 100 feet per month for tunnel-heading. These rates of progress are taken the same whether the headings be driven from deei) or shallow shafts or from a portal. By omitting shaft No. 6, the upper level has the advantage in time of excavation, which it retains, if shafts Nos. 7 and 2 be omitted. In considering how fa.r the omission of these shafts will delay the execution of the work, the experiences at other works of this kind are interesting. Mr. H. D. Whitcomb writes me as follows, relating to the Great Bend Tunnel : “Ido not know that I found any difference in penetrating from shafts or portals. So long as the hoisting-power is sufficient, there can be no material difference.” (He states very little water was found.) The earlier progress of the Hoosac Tunnel affords no criterion of the progress to be here expected, and we can only use the results from the time Messrs. Shanty commenced operations in 1869. The central shaft was already excavated to a d(q)th of 583 feet, leaving 445 feet to grade. Work commenced here May 20, 1869, and the excavation reached grade August 12, 1870, in little less than fifteen mouths, the rate being 30.1 feet per month. When it is considered that this progress was in the lower half of a shaft 1,028 feet deep, it may well be assumed that the excavation will be much more rapid in the shafts of the Alleghany Tunnel, especially in those shafts less than 300 feet in depth. In the Hoosac Tunnel, 1869, the average monthly progress from portal-head- ings was 137.7 feet; in 1870, 119.3 feet; in 1871, 130 feet; in ls72, 133 feet; and in 1873, 135 feet. Two and four-tenths months were employed, after the shaft was sunk, in prepara- tions for the tunnel-extension, such as trimming the sides of the shafts, repairing and strengthening the timbers and girders, in the removal of the pipes and pumping-ma- chinery provided by the State which were used while making the shaft, and the re- placing them by mains and pumps of larger capacity. Much of this time might havm been saved had the original preparations been adequate. The shaft, as has been stated^ reached grade August 12, 1870. Owing to the above delays and the breaking of ma- chinery, only 60 feet advance was made at one heading from this shaft, and 87 feet at the other heading, to January 1, 1871. During the next year only 277 feet were driven from one of these headings, and 153 feet from the other. In 1872, better progress was made, at one heading 1,226 feet being driven, (in Ilf mouths,) while but 119 feet ad- vance was made at the other. “ The advance of this heading was suspended during more than ten months of the year, by reason of enforced delays arising out of the large volume of water encountered, and an apprehension of developing a further increase of 698 KEPOBT OF THE CHIEF OP ENGINEERS. quantity, which would exceed the resources of the pumping-machinery provided for its removal.'’'^ It will be seen that the progress at one heading was a little more than 100 feet per month, and it probably would have been as rapid at the other, had it not been for the fear of water, or had sufficient pumping-machinery been provided. Even the former shows much less progress than the portal-headings. Shafts 2 and 6 are more difficult of access than Nos. 1 and 5, or 5 his, and preparations could not be made at the former as soon as at the latter. Eeferring now to the profile of line A, (sheet No. 1, summit division,) we see that by our original assumption the headings between shafts Nos. 5 and 6 will meet at a point 1,343 feet from the latter, and shaft No. 6 will be used for the excavation of 2,686 feet of tunnel to this time. If shaft No. 5 his be used, instead of No. 5, the headings will meet 93 feet from shaft No. 6, and only 1,866 feet will be excavated by means of this latter shaft up to this time. The time saved by the use of shaft No. 6, as computed, is nine and one-fourth mouths, but the practical saving would undoubtedly be less than this, on account of the greater accessibility of the shorter shaft and greater liability to accidental contingencies of the longer one. Merely as a matter of opinion, I would say the real saving of time would be about 6 months. We will assume 6 months’ saving of time. Shafts No. 5 his and 7 of line B differ little in their circumstances from the corresponding shafts of line A, the former being 20 feet less in depth, the latter 40 greater. The computed difference in time of exca- vation is months in favor of line B. We will call it 2 months, omitting the as the advantage which shaft No. 7 of Bne A has over shaft 7 of line B. Line A would then have but 4 months’ advantage over line B in time of execution in the tunueling of Kate’s Mountain. Assuming equal progress at all shafts and all headings, the head- ings between shafts No. 1 and No. 2 of line A would meet at a point 683 feet from shaft No. 2, and 1,366 feet of tunnel up to this time be excavated through this latter. The time saved by this shaft is computed at 6.83 months. If we suppose the same decrease of time saved in practice as was assumed for shaft No. 6, liue A will present no advan- tage whatever in the time of tunneling Lewis Mountain over line B, even if we sup- pose shaft 2 to be used in the former and omitted in the latter. The shore tunnel in Howard’s Creek Valley is 56 feet longer for the lower than for the higher level. The final determination of the cross-section of the tunnel should now be made. If passing-places be decided upon, that fact should be taken into consideration. That the towing should be by submerged chain is assumed. The boats should then follow the axis of the tunnel, and the widths of turn-outs should be sufficient to admit of boats passing easily. The width of boats is taken at 23 feet 6 inches. For the main tunnel section, the printed form, as suggested by Mr. Latrobe, is the best, affording good heights, with least excavation, besides being the shape which the excavation will naturally assume. Our estimates for the power necessary for the tow-boat have been made for a sup- posed water-way 31 feet 6 inches in width at the surface and 11 feet deep. With these data, the adjoined diagram of section at turn-out is constructed. * Report of standing committee on the Troy and Greenfield Railroad and Hoosac Tunnel for 1872. APPENDIX V. 699 The height of the crown of the main tunnel is assumed 24 feet. One-half the width of the main tunnel section is 15'. 9" One-half the width of boat is IP.G''' Clear space between boats 1'.9" Width of boat in recess 23'.6" Wooden fenders built in the w^all of recess (to preserve boats from injury by the waves) projecting 0'.3" Total width (water-surface) at recess 53'. 0" This is, certainly, as little width as can be given with the assumed widths of boat and of main tuunel. Cost per foot of ordinary tuunel — Brick $56. 66666 Stone 12. 21629 Excavation 182. 77777 251. 66074 Excess of recess over ordinary tunnel Cost per foot of recess — Brick $79 20 Stone 13 04 Excavation 242 95 435 19 251 66 183 53 This section is taken as the basis for estimates. The lower level exceeds in cost the upper level by the following items : 1,591 feet of tunnel, at $251|- per foot $400, 401 67 252.415.82 cubic yards of rock-excavation, at $1 per yard 252, 415 82 7.5,159.74 cubic yards of loose rock-excavation, at 50 cents per yard 37,579 87 1,140 feet of shafting, (10 by 40 by 1,140,) 16,888,889 cubic yards, at $10 per yard 168, 888 89 358.82 cubic yards of dam-masonry, at $8 per yard 2, 879 81 Total - 862, 157 06 This is partially offset by the following items, in which the upper level exceeds the lower in cost : 9,714.17 cubic yards embankment, at 25 cents $2, 428 54 8,635.35 cubic yards excavation, at 30 cents 2,590 60 12,888.90 cubic yards lock-masonry, at $9 116,000 10 Lumber for 7 locks, at $3,400 23, 800 00 Gates and miter-sills, at $3,150 22, 050 00 Total 166, 869 24 The cost of the lower level we see to be nearly or quite $700,000 in excess of the higher, while the saving in the time of excavation will be, at the outside, but 6:^ months. Considering the cost of the entire line,’ I do not think the lower level to possess suffi- cient advantages to balance this excess of cost, and the level of 1,713 feet above tide is therefore recommended. If the boats be passed through the tuunel in fleets and turn-outs be adopted, the times of departure should be at equal intervals. The proper intervals between the departures of successive fleets should be established before determining the number and dimensions of the turn-outs. The convenience of commerce requires that it be as short as possible, while the minimum expenditure of power requires that fleets be as large as can be handled easily. We will suppose the interval to be two hours; assum- ing eight lockages in this time in each direetiou, the fleets will consist of eight boats. Two fleets will then emerge each two hours, (oue from each portal.) In the formula n=:Nv, N will then equal unity; n will conform to v. For economy of power the rate of movement should be slow, while economy of construction requires that it should be rapid. The following table (IV) is formed by reducing the velocities in Table I to 1 mile per hour, assuming the resistances to vary as the squares of the velocities, and taking the supposition as correct that the resistance for the first boat per unit of submerged section is double that per unit for the succeeding boats of the same fleet. Experiment 6 is omitted, as the fleet w^as formed of rafts. 700 REPORT OF THE CHIEF OF ENGINEERS. Table IV. Xumber of exj erimeat. ! -g I 1 2 3 4 5 |1 1 1 1 1 1 2 ‘2 2 2 2 2 2 2 A B C I 64. 04 128. 08 SCO. 1 .s;-^8. 5 400.6 ‘'< 2. 12 • 68. 89 57. 05 144. 24 104. 83 580.4 519.3 .564. 8 G.)8. 5 599.6 K L I 65. 66 73. 73 05 760.6 693. 7 M X 0 ^ 69. 96 87. 19 227. 11 ( 634. 4 < 793. 6. (. 656. 2 3. 905 4. 665 3. 8-22 4.024 3. 600 2. 899 3. 123 3. 077 3. 709 3. 386 2. 793 3. 494 2. 889 1^3 T ^ 3 U I 71. 04 130. 78 272. 86 1, 048. 6 1,021.5 3. 843 3. 780 3 V 3 W 3 X I 72. 66 135. 63 280. 95 1, 036. 8 1,171.9 1, 151. 6 3. 699 4. 171 4. 099 74. 47 401. 50 550. 44 1,906.' 3. 464 The arrangement of this table has made apparent a fact which had escaped atten- tion before, viz, that in those experiments in which three results are given, the second shows invariably a greater resistance than either of the others. The direction of movement of the different boats or fleets was the same in all, and the second result refers to the middle of the tunnel, where the resistance was then greater than at either end. In the second, fourth, and seventh experiments, the first velocity and resistance noted were near the middle of the tunnel, the second nearer the farther end, thus confirming the indications of the first, third, fifth, and eighth experiments. In the ninth experiment the space passed over was from near the middle to near the end. It is necessary, therefore, in discussing ihe results of sucli experiments, to con- sider this fact, and not to compare the data furnished from different portions of the tunnel with each other. It is to he inferred that if the tunnel were longer, the resistance near the middle would he still greater than here noted. Considering the small number of experiments, this table demonstrates in a remark- able manner the deduction of M. Bazin from Table I, which is, that the co-efficient of resistance for each boat following the first in a fleet may with safety be taken as one- half this co-efficient for the first boat. The third part of the eighth experiment is-apparently an excention to this rule, but in this part the point of departure was 640 feet nearer the middle of the tunnel than the point of departure for the third part of the first experiment, though the terminus was the same in each case. An inspection of Table IV would seem to show that in our calculations for the power necessary for our tow-boat, we may assume 5 pounds resistance per unit of total sub- merged area [=(8-{-l) (23'.5x6'.5j =1374.75,] and the total resistance for 1 mile per hour =6873.75. With the proposed area of water-section, I estimate that the current through the tunnel would be one-third mile per hour to feed the canal east of the summit-level. If we suppose that one-third mile per hour will be necessary, in addi- tion, to overcome this current, there will be required 1*2,220 pounds total traction. Labrousse deduces the following formula for the useful elfect of trac'ion with sub- merged chains : V 255 H represents the depth of water in meters, but is really the vertical distance from the bottom to the point where the chain leaves the pulley. We will be safe to as- APPENDIX V. 701 -sume this distance at 5 meters. The tension of the chain as it leaves the pulley (which is the resistance to be overcome by the machine neglecting its own friction) is then 12,338 pounds. The work of the machine at this point will be, say, 44.1 horse power. In these estimates the resistance of the tow-boat itself has been neglected, but we have considered the westward-bound boats alone, and supposed each to be loaded to 6 feet 6 inches. It is probable that the draught of even heavily-loaded boats will not exceed 6 feet 4 inches, and many of the w^estward boats will be lightly loaded. I think two-thirds of the above estimate of total submerged area are to be sufficient for westward fleets. The actual velocity of 1 mile per hour being obtained, one turn- out per mile will be necessary. At times, if sufficient water flows in Brush Creek after heavy rains to supply tempo- rarily the quantity necessary for the canal to Dunlap’s Creek, (in which the water will aLso t)e high,) there will be no current in the tunnel, and the formula a = N (v n t) must be satisfied. Substituting for n and N their values, we have v t = 1; that is, the time of passing from one turn-out to another, (being one mile apart,) added to the time lost at a turn-out, must not exceed one hour. We have calculated the resistance for 1-J- miles per hour; v will then be f, and t must not exceed ^ of an hour, or 15 min- utes. This would seem ample time for the maneuver at the turn-out in still water. The next thing to be considered is whether the strain may not exceed the strength of a chain of ordinary dimensions. If w^e neglect the friction on the bottom, the greatest strain occurs when all the fleets moving in the same direction are alone using the chain. This will happen when the eastward fleets are all in the turn-outs at once, which may frequently occur. We must then provide sufficient strength for 5 fleets. A chain employed on the Upper Seine* is made from 0.886-inch iron, and sustained a test-strain of 26,443 pounds before use. This chain weighs about 7.6 jDounds per run- ning foot. Lagrend assumes the friction equal to one-half the w^eight of the chain; about one- seventh must be deducted for loss of weight in water. The friction would then be more than 3:1 pounds per running foot, aggregating 16,926 pounds per mile. The dis- tance between fleets would be 2 miles, and, therefore, the tension would become zero before half this distance were reached, as we have found the tension for each fleet to be 12,220 pounds. Labrousse mentions an instance in which, if we suppose a force of traction of 2.2 pounds necessary on an open canal for each ton with a velocity of 3.28 feet per second, the friction must have been 0.58 of the weight of the chain on the bottom. Or, sup- posing this force of traction but 1.1 pounds, the friction would be 0.29 of this weight. Under this latter supposition the resistance would aggregate 20,365 pounds in the dis- tance between 2 fleets, and the tension would become zero in about 6,332 feet ahead of each tow-boat. It is, then, more than probable that the tension necessary on the chain will at no time exceed the tension necessary for one fleet, which we have seen is less than 50 per cent, of the test-strain of the chain in use on the Seine. The strength of the chain is then an assured matter, assuming the net velocity of movement at one mile per hour with 8 boats iu tow. Assuming four lockages per hour as the maximum trade to be provided for by turn-outs, we will now compare the 2-honr interval with others, viz, 4, 3, 1^, and 1 hour intervals, supposing the tow-boats in each case to work to 44.1 horse-power, (effective.) Table V shows this comparison. * Lagrene. TAliLE 702 REPORT OF THE CHIEF OF ENGINEERS, •inaniaSnejjB ^so(Ib91[o J3AO :jsoo jo Bsaoxg; $1, 023, 034 32 438, 452 28 97,421 84 69, 752 34 •uopo9s-{onmi; AjBuipjo JteAO i^soo JO ssaoxg; Cf CO ^ Cl CO a ^ T-HOC5Cil^ 1 -H o c* iTi a Oi Ci ^ rr a: GO — < ^ 00 c: lo CO o ^ 00 O lO Ci •J99J TIT ‘S9 -SS909J JO qjSao’t ojBSojgSy 13464 10296 8448 8298 7920 'oraij ono jb X9nnnj uiqjTAV s’j9iaB9js JO jaqran^ C5 o CO T-l r-* (•{9A9{ joddn) ‘[onanj mqjTAi sjno-aanj jo jgqtunj^ CO O OC c: (M •J9J -n99 OJ J9JU9D TnOJJ ‘J99J TIT ‘fe-9SS999A aaoAvioq siBAjaj'ni O O ic Tf ^ ^ CO ^ OO (N -rr tt CO cc ir: rr CO •Agnnd S9 ab 9^ qi sb ‘ spuTTod u[ ‘aiBqo jo uoisaojr — • O'. 00 r-* lo CO — CC O crs CO O CO -H irf ct ^ o' •qn9.i.Tuo 9rao9J9AO oq Aans -89090 aiioq .i9d 9[iai paiqq -900 SoTsoddos ‘AqTooqoA cc O O CO t- cr- o o CO 'TP X' O C't o ^ ^ C't •jnoq jod ‘89X110 OT ‘S9iqtOO[9A pO'j’udTOOQ l'- ^ CO CO X ^ Cl CO ir: i.o 00 -Xj CO LO O’? O T-' CO o •qno -o.inq JO qq“o9x oiBqqo oq ooBds 09.TJ joj qqa9q-9oo ppY cc O CJ o TJ* — lO On* CO C* t- O O CO — .-1 •qgaj ord ^diTOOo oq jBoq-JAoq pn« jBoq qoB9 SoTSodaus ‘qggp jo iiq^fogp; o o o o o CO cc •'T O O LO C; X CO Cn* T-i •sq99p OT ejBoq jo jgqoinx i ^ S'* 00 'o j i cS a s- ® o p fc- r ® ® ? c a P^HE^CO { APPENDIX V. 703 It is thus seen that for this power of tow-hoat, economy of construction is on the side of small fleets or short intervals. The 3 and 4 hour intervals are rejected. We can provide for the assumed trade with this expenditure of power by either of the three latter intervals, and with little difference of cost; for the 2 hour interval will require 4 less steamers, and the 14-hour interval 3 less steamers than the one-honr interval. The original cost of these steamers, with wages of crews, running expenses, and repairs, will go far to remove the inequality. It may be interesting to note the phase the question takes by assuming the same tension of chains for the three latter intervals; the tension assumed is that for the 2-hour interval, 12.33::! pounds: InterA’als of departure. Computed velocities, miles per hour. Net velocities, miles per hour. Etfective horse-power necessary. Intervals between re- cesses from center to center, in feet. Number of recesses within tunnel. Aggregate length of recesses. Excess of cost over ordinary tunnel-sec- tion. Two 1. 3333 1.000 44.4 5280 8 8448 $1, 558, 909 44 One and a half 1.5114 1. 1782 40.6 4666 9 82'l8 1, 531, 229 94 One 1. 7883 1. 4550 58.7 3832 11 7260 1, 339, 087 80 Under this supposition, the 2-honr and 1^-hour intervals remain the same in regard to cost. The cost in the 1-hour interval is reduced $122,389.80, the number of steamers necessary is reduced by 1, but the power of each tow-boat is increased 14.1) horse-power. If we suppose the friction one-half the weight of the chain on the bottom, it would amount in the distance between fleets to 13,123 pounds with the chain in use on the Seine, being but 765 pounds more than the tension assumed. Deducting the amount of chain raised from the bottom by the tow-boat, these two forces would be about the same. But if the friction should be less than one-half the weight, the surplus tension would be transmitted forward, and the accumulation from the different fleets might be more than a proper chain should be subject to. High velocities with short intervals should on this account be avoided. Low velocities with short intervals effect no sav- ing in cost, while rendering ventilation more difficult because of the greater number of steamers required. All things considered, the 1^-honr interval with the lower power is recommended as the arrangement to be adopted in the method by fleets and passing-places. It requires 1 more steamer than the 2-hour interval, but a higher velocity can be more safely and easily obtained should it become necessary from any cause; and the saving of time will amount to a good deal in a yeaPs trade, sufficient to provide for the greater diffi- culty of ventilation, should a difficulty of this kind arise. By spacing the turn-outs at distances of 4,435 feet, as required by Table V, beginning at the eastern portal, the ninth turn-out will be about 1,700 feet inside the west portal, the tenth will be between the first and* second crossings of Howard’s Creek, and the eleventh will be within the short tunnel. The proper place for this last turn-out, being the first approaching from the west, is below the short tunnel, where the boats will naturally collect and form tows. If this turn-ont be placed here, and eleven turn-outs be used, one of two alternatives must be adopted: either the turn-outs must be placed farther apart than 4,435 feet throughout the line, or one turn-out must be located be- tween the main and short tunnels, and the open cut be made sufficiently wide to de- crease the resistance, so that sufficient speed may be obtained here to make up for the greater distance between turn-outs. This first is not advisable. The second is more expensive than the excavation of an additional turn-out in the open cut. Accordingly the first turn-out from the west, which is bat a harbor for the formation of fleets, is placed below the short tunnel ; the third is immediately outside the west portal of the main tunnel. Nine turn-outs will be necessary within the long tunnel. There is suffi- cient available space in a straight line outside the east portal for the formation of a fleet of six boats, still it is desirable that it should not be necessary for the fleet just formed to enter immediately on the exit of the eastward boats. It is well to have a margin of time to remove the eastward boats out of the way; to attach the westward tow-boat to the chain, and to assure a proper condition of its fires before entering. To provide this margin of time,the first turn-out from the east portal should be at a less ditance from the portal than the distance between turn-outs. By spacing the turn-outs at distances of 4,350 feet from center, to center, which is 85 feet less than required by Table V, beginning at the turn-out just outside the west portal, the conditions seem to be fulfilled in a reasonable manner. The second turn-out in Howard’s Creek Valley is located above the aqueduct for pass- 704 EEPORT OF THE CHIEF OF ENGINEERS. ino’ the creek over tlie cut and on the side opposite the'creek. The creek is to be turned iuto the sloughs about opposite the portal, and where necessary the earth excavated from the tunnel-approach will be formed into a dike to prevent floods from flowing into the cut. The cut is to be arched over for a distance of 1,200 feet above the short tunnel, to pass Howard’s Creek and its floods. The dike will terminate at the upper extremity of this arching. The connection of the creek and canal below the short tunnel is to be made as follows: The creek-bed will be excavated for 600 feet up- stream to a width of 80 feet, the lower half to the level of canal-bottom, the upper half to the level of the comb of the dam below, viz, 1,721. Across the lower end of the upper half wall be a dam 5 feet high, wdiich may be natural if the rock be met with and will permit it. This dam will have a gate, which may be opened at low water, permitting the artificial pool above to be drained off, when the sand and gravel brought down by the creek and there deposited may be removed. A supporting-wall for the movable materials for the creek-bed above will probably be necessary to keep them out of the catch-gravel pool. The Howard’s Creek dam is located below Caldwell Station, where the valley widens out as it approaches the Greenbrier. This location, though requiring a dam about 40 feet in height, seems to be the best. I at first thought of adopting a location near station 67 71.2, which would require a dam about 15 feet high. But this would ren- der necessary the constructing of more than 1,200 feet of feeder over very bad ground. In leaving this dam one of two methods might be adopted : 1st. A guard-lock at the dam on the left of the valley ; then a cut not less than 30 feet deep for a quarter of a mile, and an aqueduct over Howard’s Creek. 2d. Lock down from dam on the left and enter, near the mill, the pool of a dam located at the adopted site. Either of these, with the extra feeder-canal, would undoubtedly be more expensive than that recommended. A flight of two locks is necessary at the dam. The canal joins the Greenbrier division in the pool of a dam, description of which will be givt u in the report of that division. The lock at the Greenbrier is 14 feet lift. There will be required east of the summit the following amount of water: Cubic feet per minute. Evaporation and filtration, sixteen miles, 200 cubic feet per mile per minute.. 3, 200 Lockage, (supposing six lockages per hour and one lock-full toeach,)=:[14 X 24 X 120] X 120 equal 241,920 cubic feet per hour =4, 032 Waste at structures = 500 Leakage at gates =2, Total 9,732 IMcNeil found the minimum flow of Dunlap’s Creek to be 9.43 cubic feet per second, = 565.8 cubic feet per minute. The tunnel at the time of minimum flow^ would be required to deliver 973 ,* — 565.8 = 9153.2 cubic feet per minute. The area of its water- W'ay is 338.25 square feet, and the velocity, therefore, sliould be 27.1 feet per minute, say i mile per hour, (=29^ feet per minute, delivering 9,709 cubic feet per minute.) Tlie distance from Howard’s Creek dam to Inrush Creek is somewhat short of 60,000 feet. The fall necessary in this distance for a velocity of 30 feet per minute is ^ of a loot, nearly. To provide lor this fall, and also to correct in a degree any irregularity of feeding, the comb of the dam in Howard’s Creek is established 8 feet above canal- bottom — i. e., at reference, (1,721.) The comb of the dam in Brush Creek Valley is to be one foot higher than this, to prevent excessive w^aste of water over it b}^ the waves formed by the eastward fleets. The top of the lift- wall of the first lock is on a level with the bottom of the tunnel, so that navigation may not cease as long as there be sufficient wat r therein to float a brat. This will still further tend to correct irregu- larities of feeding. The conformation of Brush Creek Valley requires that the locks for entering and leaving the basin should be on the hill-side to the right of the dam, while the canal should be on the left of the creek, below station 77-f-50. Accordingly, a dam is located near station 77, from the pool of w^hich the boats leave by a flight of three locks. To prevent waste of water, the comb of this dam is 8 feet above canal-bottom. The flights of locks at both dams are double. The lift of all locks in Brush Creak Val- ley is 8 feet. The levels are necessaiily short, and the ov'erfalls should be so arranged as to retain rather more than the necessary depth of 7 feet. Through a good pjirt of this distance, a dike wall be necessary to restrain the floods of Brush Creek. This dike, also serving the purpose of embankment to the canal, should be several leet above the creek-bed. For this reason, the excavation in this valley is generally of little depth, wTiich permits the levels to oe wider than the ordinary canal, making up somew^hat for their little length. The connection with Lorraine’s location is made in a pool in Dunlap’s Creek. I APPENDIX V. 705 would recommend the pool of this dam, as well as the others iu Duulap’s Creek, to be 8 feet above canal-bottom, to prevent waste of water aud to correct irregularities of feeding. A riprap protection is estimated for on the upper part of the dike, where the channel-way of the creek is somewhat restricted in width. Below the last lock, a loose-stone breakwater is provided on the up-stream side, to protect the boats at the tail of the lock from the current iu Dunlap’s Creek during high water. The feeding arrangements are next to be considered. The maximum lift of lock east of the summit is 14 feet, which is also the lift of the lock connecting the summit with the Greenbrier division. Taking this lift as the unit of lockage, supposing 6 lockages per hour, each requiring two locks full of water, there will be required each hour for lockage, (2 X 6) (14 X 24 X 120) = 483,820 cubic feet, equal. 8, 0(54 cub. ft. per min. Evaporation and filtration, 28 miles, at 200 cubic feet per mile, 5,600 cub. ft. per min. For leakage at lock-gates 4,000 cub. ft. per min. Waste at structures 1,000 cub. ft. per min. ’Total 18,664 cub. ft. per min. Say, 320 cubic feet per second as the quantity to be delivered at the summit by the feeder, if we neglect the flow of Howard’s and Dunlap’s Creeks. Mr. Hutton’s proposed location for feeder-dam for the (1,700) level is near the mill- dam above dam A, (Sheet No. 4, summit division.) From there the feeder-canal was proposed to keep to the hill-side on the left of the valley, aud to pass round the point near the Greenbrier bridge, uniting with the canal below station 106 76 of the How- ard’s Creek line. (Survey of 1874.) Mr. Latrobe suggested the adoption of a feeder-tunnel through the point about opposite station 115 + 05 of Howard’s Creek line. An examination with this view of saving feeder-canal by tunneling the point, led to the discovery of the cut-off by the ravine from the Greenbrier, between dams Nos. 1 and B. A tunuel here 1,340 feet in length will save at least 7,600 feet of feeder on very bad ground, besides 7.6 feet in height of feeder-clam. The location for a feeder-dam at the bluff, near dam B,is not available for the 1,713 summit at least. For this level the dam would be about 50 feet in height, with a guard-bank more than 1,500 feet in length, the river-end of which would be about 60 feet high. The pool would be of no use as a reservoir, for, of course, the water could not be drawn off below the level it is intended to supply. The location near station 108 is recommended. No guard-bank will be required and no valuable ground will be overflowed. The normal level of Die water-surface on the summit is 1,720 feet. This level is taken as the bottom of the feeder, where it discharges into the Howard’s Creek pool, and the slope of the bottom is assumed at one foot in one thousand feet. A rectangular section (18 X 5 = 90 square feet) is adopted as the standard. This with a slope of i^fo-cr wili give a velocity a little in excess of 4 feet per second, deliv- ering 360 cubic feet in that time. The sections of the feeder-tunnel at the Greenbrier end are shown by the figure annexed. The arch is horizontal and the bottom falls for the necessary slope. A dam is pro- posed across the ravine at the Greenbrier end of this tunnel, which will provide a waste-weir at this point. The bottom of the tunnel at this end will be 1,721.5 above tide. Five feet of water will make the surface-elevation 1,726.5. The top of the waste-weir is established at 1,727, aud the feeder-bottom at station 24 is taken 1,722. The dam should have a gate so arranged that the pool may be drained, and the sedi- ment brought down by the feeder washed into the river. The excavation to the ravine above dam A will be mostly in red shale. The cross- section at station 24 may be taken as the type of this portion. It is recommended that a wall be built on the river-side 7 feet in height (see cross-section station 21) and 51^ feet in width, resting on a foundation of concrete 2 feet thick, and that the exca- 45 E 706 . REPORT OF THE CHIEF OF ENGINEERS. vation be made 18 feet in width on the bottom. The bine line from waste- weir to feeder- dam (Sheet No. 4, summit division) is the outer edge of feeder-bottom. Eleven culverts will be required on the feeder-line. The wall on the river-side of the feeder ends at station 72 50. Above this for some distance embankment is considered preferable in view of expense. To prevent excess- ive waste of water from the feeder the following is recommended : The excavation to be made 1 foot below feeder-bottom, then about 9 inches of puddle-clay carefully laid on this, about 2^ inches of stone, broken small, to be laid and lightly rolled, then a layer of hydraulic mortar to be thrown on and rammed carefully, so as not to disturb the broken stone. I should fear that puddling alone would be washed away by the cur- rent. The inner slope and outer slope (when embanked) to be paved to 7 feet above bottom. Paving to be 6 inches thick, laid in mortar. ANALYSIS OF COST PER RUNNING FOOT. One-sixth yard broken stone, at$l |;0. 17 One-half yard puddling 0.25 One-quarter barrel cement 0. 40 Two cubic feet sand 0. 05 Laying broken stone 0. 03 Laying and mixing mortar 0. 10 Total per running foot 1. 00 Paving per running foot 1.50 The length of canal to be filled from the summit feeder or feeders is, in round num- bers, 150,000 feet. In this length are eight pools in Dunlap’s Creek, two in Brush Creek, and one in Howard’s Creek, and the canal in Brush Creek has a section larger than the standard adopted. The tunnel water-way has less sectional area than the canal sec- tion. Supposing the average throughout the line to be the canal section = 441 square feet, there will be required to fill the summit division 66,150,000 cubic feet. If we suppose while filling the canal the losses by leakage at gates, evaporation, and filtration to equal the amount supposed when estimating for permanent expense, 160 cubic feet per second would be required to supply these losses. The feeder will deliver 360 cubic feet per second. If we neglect the flow of Howard’s and Dunlap’s Creeks, 200 cubic feet per second of the feeder-delivery will be the available net amount for filling the trunk of the canal. The time required would be somewhat less than four days. This would not seem too long, as some time is required to permit the boats to approach from below. If artificial ventilation should at any time be necessary, the simplest method would be to utilize the summit-supply of water to provide the power. For the upper level, the Greenbrier feeder is not available for this purpose, except at the great expense necessary to build a longer feeder. This feeder would be available for the lower level through a fall of 12 to 13 feet, but the power would be about two miles from the por- tal, and this extra length of pipes would be necessary.* Lorraine made surveys for and located reservoirs on Howard’s and Jerrico Creeks, and on Tuckahoe Creek, and presents the following tablet of the contents of each: Reservoir. Area in acres. Contents in cu- bic yards. Drainage area in sq. miles. Cubic yards of rain-fall, 37 inches. 40 per cent, for drainage. Howard’s and Jerrico 248 159 9, 276,410 7, 693, 464 6, 750 9, 522 33, 573, 870 47, 361, 539 13 429, 548 18, 944,616 Tuckahoe The Howard’s and Jerrico reservoir contains 250,463,070 cubic feet, which would sup- ply 581 cubic feet per minute = 36,512 foot-pounds per foot of fall. We can get 60 feet of fall at the west portal, which from this reservoir would give 2,190,720 foot- pounds per minute ; 75 per cent, of this would be 1,643,640 foot-pounds per minute, nearly 50 horse-power. The Tuckahoe reservoir contains 207,723,528 cubic feet, which would supply 480 cubic * This availability of the Greenbrier feeder is an advantage possessed by the lower level not mentioned heretofore. tl am indebted to Mr. J. M. Harris, superintendent James River and Kanawha Com- pany, for this table. APPENDIX V. 707 feet per minute for 300 days = 30,000 foot-pounds per minute for each foot of fall. The site of this reservoir is now cut up by railroad, but more than one-half the above con- tents could probably be counted upon — 15,000 pounds per minute through a fall of 60 feet would give 900,000 foot-pounds per minute, 75 per cent, of which would be 20^ horse-power nearly. It would not always be necessary to draw from the coutents of these reservoirs, as on many days the flow of Howard’s Creek would supply all the power necessary. Some of the water from these reservoirs would be lost by evapora- tion, &c., before arriving at the portal. But the natural drainage pertaining to them will su])ply much more than double the quantity supposed to be used. It will be observed that while arranging the tunnel for four lockages each way per hour, the estimate for the water-supply has been made on a basis of three lockages each way per hour. If there be but three lockages on an average, it would probably occur frequently, as the fleets must start promptly on time, that a boat slightly behind time for one fl^et would with rapid lockages in the next interval form a fleet of six boats, and those fleets which move the slowest regulate the rate of progress through the tunnel. Then, as now arranged, the maximum lift of lock is 14 feet. Six lock- fuls would most probably be sufficient for this lift, as boats cannot be so readily passed as through locks of less lift, and there must be some alternate passages. If eight lockfnls be really necessary, the amount of water to supply the summit division, the losses being as supposed, will be 356 cubic feet per second, which is less than the delivery of the feeder. We have, besides, the flow of Howard’s and Dunlap’s Creeks. If required, the depth of water in the feeder may be made more than 5 feet, thus increasing the delivery. The velocity in the tunnel would be increased. If eight lockfuls be necessary, the amount required east of the summit would be 11,076 cubic feet per minute. Deducting the minimum flow of Dunlap’s Creek, the remainder, 10,511 cubic feet per minute, would be the amount to be delivered by the tunnel. The velocity would be 31.1 feet per minute, (= 1,866 feet per hour, say miles = 1,920 feet.) The necessary horse-power for tow-bo;»t won.d be 44.72 horse-power, an increase of but 0.62 horse power over the amount estimated for. I would recommend that the lifts of the locks in Dunlap’s Creek Valley be changed to decrease the maximum lift. If 8 feet should be adopted throughout the line, the quantity of water required east of the sutnmit, supposing eight lockfuls to be neces- sary per hour, would be decreased 2,304 culnc feet per minute. The amount to be de- livered by the tunnel would b-j 8,207 cubic feet per minute, and the power necessary for the tow-boat would be 37^ horse. If a reservoir were constructed on Dunlap’s Creek of sufficient capacity with the flow of the creek to supply the losses from evaporation and filtration below the mouth of Brush Creek, the amount to be delivered by the tunnel would be still further reduced by 2,822 cubic feet per minute, and the power necessary for the tow-boat would be only 31|^ horse. This reduction of power would materially diminish the con- sumption of fuel in the tunnel, and to that extent relieve the ventilation, besides de- ci easing the running expenses and first cost. The reservoir necessary would not be large. McNeil found the average flow of Dun- lap’s Creek, from August 7 to September 27, 1826, to be 19 cubic feet per second; and from the 17 th of May to June 16, 1827, the average was 94.41 cubic feet per second. The amount required would bo less than 50 cubic feet per second. This is a matter to which attention should be directed while the tunnel is being excavated. Lorraine made a survey for a reservoir on Cove Creek, and it is one of those shown on his map of the summit, but I cannot learn that he made an estimate of cost or left any record of its capacity or area of drainage. During the spring, fall, and winter seasons there is no doubt that Dunlap’s Creek will provide the necessary supply. These being the times of greatest trade, it may be that no reservoir will be required, and that the power necessary, measured on the chain, will be at all times less than 40 horse. If the com- merce of the line should decrease during slack trade to six lockages per hour, the inter- vals between fleets can be increased to two hours, when the eflective power necessary will be 27.3 horse. In closing this report I would make some remarks on the apparently great difificul- ties attaching to the construction of the summit tunnel. A tunnel of the length proposed (nearly eight miles) is no new or experimental mat- ter at all. The Mount Cenis tunnel is nearly as long as this, while the Saint Gothard tunnel through the Alps, now being excavated, will be about 8,000 feet longer. At neither of these tunnels could shafts be found to expedite the construction, while the Alleghany tunnel will have ten or twelve of moderate depth, thus dividing the long tunnel practically into a number of shorter ones. The deepest shaft is 400 feet less depth than the central shaft of the Hoosac tunnel. The rates of progress at shafts and at headings are assumed at 30 feet per month for the former and 100 feet per month for the latter. The progress at the central shaft of the Hoosac tunnel, as mentioned, was 30.1 feet per month, while the excavation was carried on in the lower half of a shaft 1,028 feet deep, at which insufficient preparations had been made for the proper construction of the tunnel. 708 REPORT OF THE CHIEF OF ENGINEERS. Mr. li. D. Whitcomb writes me as follows : “ The 370-foot shaft at Great Bend tunnel is the best guide I have to offer for progress in a deep shaft. It was reinoved by con- tract to a depth of 70 feet. The remaining 300 feet was sunk in six months. Some of the rock was a very hard sandstone. I think the best progress made was about 70 feet in one month.” I have not the slightest doubt the Alleghany shafts could be sunk at the average rate of 50 feet per month easily. Mr. Whitcomb informs mo that the greatest progress he remembers (in heading) at Great Bend was 180 feet per month. “In very hard sandstone in that tunnel we made 80 feet. A fair average would be 125 feet. In the Lewis tunnel, I suppose 65 feet would be a fair average without machinery, i. e., with- out machine-drills.” Subsequently he wrote: “Mr. Talcott says he thinks I over- estimated the average progress at headings, i e., counting all delays. He says : ‘An average would be 100 feet per month. But I will add that the contractor was inexx^e- rienced, and I feel sure that another such tuunel would be driven faster.’ ” The Great Bend tunnel was, I believe, excavated by hand-drills. The progress at the Hoosac tuunel was about 130 feet at x^ortal headings, and would have been more than 100 feet from the deep shaft if sufScient pumping-machinery had been provided. At the Mount Cenis tuunel the average monthly progress at headings, from 1864 to 1870, was as follows: 1864 1865 1866 Feet. 148.6 1867 168.3 1868 140.1 1870 Feet. 206.5 180.4 223.5 For the entire seven years the average monthly progress was 180 feet at a heading. The Saint Gothard tunnel through the Alps is the last great enterprise of the kind 'put in execution, and will be 49,733 feet in length. In 1874 the average monthly prog- ress at a heading was 247.6 feet. For the first four mouths of 1875, (tlie last con- nected account I have seen,) the average daily x^rogress was more than 10 feet, and, in Sex^tember, 1875, the last rex)ort I have found, the x)rogress was 419 feet at one heading and 344 feet at the other. We have every reason to €xx)ect more rax)id x)i’ogi’ess at the Alleghany tunnel than has been assumed. Attention is particularly invited to the profile of the recommended line. The deep- est shaft is about 600 feet. If this be used, there is no doubt whatever that the western heading of Kate’s Mountain will be open to the x>ortal in two years. Twenty-five and three-fourths mouths is the estimated time, assuming x)rogress at shafts at 30 feet Xier month and 100 feet at headings. If shaft 7 his be used, Kate’s Mountain heading (western) will be open to the x^ortal in less than 20 mouths. The eastern heading of Lewis Mountain will be open to the portal in less than 11 months. At the end of the second year the work will be in the following condition: All shafts will be excavated : the longest headings will be ox^en to the x>ortals; 21,554 feet of tunnel (more than half the entire length) will be excavated, and operations will be carried on at the portals and but 6 shaft-headings. At the end of the fourth year, the entire tunnel will be excavated except about 1,000 feet at Kate’s Mountain, both head- ings of which will be open to the portals. The more the matter is considered, the less do the difficulties appear. Estimates in detail of the line from Dunlap’s Creek to the Greenbrier River, includ- ing those for the feeder, accompany this report. The total of this division, inclusive of contingencies, as shown by the recapitulation, is $16,387,757.45. The estimate for the tunnel includes arching throughout ; should any portion require no, or but x^artial, arching, the amount thus saved will fully cover any increased cost of excavation due to the harder rock. It may be that the excavation at the recess will assume the x^ointed form. If so, the cost will be somewhat increased at these points, but not materially, as the extra excavation will come out as loose rock. The cost of tunnel-excavation is taken at $5 per cubic yard. I believe this to be a fair price. Shaft- excavation is taken at $10 per cubic yard. This is but one-half the price assigned this excavation in previous estimates. I judge $10 to be sufficient, from the following information kindly furnished me by Mr. Whitcomb, viz : “ The main shafts (two) of the Great Bend tunnel were 170 and 370 feet (about) re- sxiectively. The contract-xirice was $6 x>er cubic yard. The contractor excavated the shorter shaft, and, say, 90 feet of the deeper; the company then completed the deeper one. Iffie contractor received an allowance on the shorter shaft, bringing the cost up to between $11 and $12. 1 think the deep shaft co.st us between $13 and $14 per cubic yard ; size, 8’ X 18'. We had to wagon all machinery 40 miles across a rough country or send it via Greenbrier River in boats. I suppose such a shaft would be doue now for $10 x^er cubic yard.” The greater size of the Alleghany shafts (10' X 40') will tend to reduce the price per cubic yard. APPENDIX V. 709 ' Brick arching is estimated at $12 per cubic yard, and the masonry lining below the water-surface at $8 per cubic yard. In the estimate for dam and lock masonry the cost has been taken at $9 and $10 per cubic yard, respectively, for the purpose of keeping a uniform standard with the New Eiver estimates. This course is also adopted with the Greenbrier division. As far as known of the summit and Greenbrier, some difficulty may be encountered in getting stone which can be cut easily or readily dressed to shape, and for that rea- son masonry may cost somewhat more than in the New Kiver division, where unlim- ited quantities of fine stone abound. In a large enterprise of this kind the development of resources may be expected which have been hitherto undiscovered or unemployed. This is most probable with regard to cement, as I understand that fine cement-stone exists at Callahan’s, (a short distance westward of Covington.) Proximity to the ce- ment may in part make up for the greater distance from good quarries. The estimates for masonry, embankment, and excavation for this division, exclusive of the great tunnel, were made by Lieutenant Maguire, with the exception of some changes recently made in the dams. In comparing the estimates of this line with others in the vicinity, it should be remembered that these estimates cover the entire distance from the Greenbrier at the mouth of Howard’s Creek to the mouth of Brush Creek. Also, it should be remembered in comparing the entire central water-line with any other that the higher summit of McNeill’s is still available, by which the summit may be passed at an elevation of 1,916 feet above tide, with a tunnel 2f miles in length. This is, I believe, with approach-cuts, 50 feet in depth, and of moderate length. All of which is respectfully submitted. Thoimas Turtle, Mrst Lieut, of Engineers. Major William P. Craigiiill, Corps of Engineers, U. S. A. recapitulation of estimates for summit division, survey of 1874. Total estimate of Brush Creek $827, 787 49 Total estimate of tunnel * 12,376,608 29 Total estimate of Howard’s Creek 1, 251, 162 69 Total estimate of feeder-line - 444, 134 77 Total 14, 899, 693 24 Contingencies, 10 per cent, of total 1, 489, 969 32 Grand total 16,389,662 56 report on GREENBRIER DIVISION, SURVEY OF 1874,. BY LIEUTENANT THOMAS TURTLE, CORPS OF ENGINEERS. Baltimore, Md., July 24, 1876. Major : I have the honor to submit the following report on the surveys and esti- mates for the Greenbrier division of the central water-line : After the completion of the surveys for the summit division, those for the Greenbrier commenced, Mr. R. H. Talcott being left in charge in the field, and Mr. W. S. Walker taking one of the transits. The surveys included those for a slack-water navigation and for an independent canal from the western approach of the summit tunnel to the west portal of the Great Bend tunnel of the Chesapeake and Ohio Railroad. From this point to the niouth of the Greenbrier River the surveys were made by Mr. Hut- ton’s New River parties, under the immediate charge of Mr. C. R. Boyd. From the combined notes of these surveys the maps, four in number, which accom- pany this report have been made. Sheets Nos. 1, 2, and 3 are on a scale of 1 inch to 200 feet, and show the entire valley as covered by our surveys, while sheet No. 4, on a scale of 1 inch to 600 feet, shows the entire Great Bend. In the autumn of 1874, after the completion of the surveys, preliminary estimates for the slack-water system were made, to accompany a report made at that time. The detailed report was made by Mr. R. H. Talcott. Since that time further study has modified the plans so that a new re- port will be necessary, though free use is made of that made by Mr. Talcott. No re- port had been made for the independeut^canal, as the time did not permit the making of e.stimates. The slack-water system will first receive attention. “ The survey for the slack- water navigation of the Greenbrier River was begun on the 710 EEPORT OF THE CHIEF OP ENGINEERS. 22(1 of September, at a point about two miles above the mouth of Howard’s Creek, at which it was assumed that the direct tumiel from Brush Creek to the Greenbrier River would debonch, supposino that tunnel to be the one adopted. The transit-line began at station 52-}- 11 of an offset line from the direct tunnel, and the levels were started from a bench on a maple about 200 feet to the left of that station, and the elevation assumed was 1,704.24 above mid-tide, that being the mean of two lines of levels run during the previous survey for the summit level.’”^ In my report for the summit division I recommended the valley of Howard’s Creek for the western approach of the summit, and therefore the dams located above dam No. 4 will not be necessary for this line, and no estimates have been made for them. In the arrangement now recommended the dams have been made as long as can be judiciously done, and the heights of guard-walls and guard-banks have been made to correspomi to the lengths of the dams and to the estimated discharge of the stream. For this latter we have no certain data. I am informed by Mr. Talcott that he has personal knowledge of a flood at Graham’s Ferry 20 feet above low water, and this flood was not as high by 5 or 6 feet as the highest known, according to the account of the inhabitants near there. A point of highest water was found nearly opposite Cald- well’s mill, above Greenbrier bridge, where the rise was about 12 feet. I have taken the former as the safe gnide. Herewith is a profile of the Greenbrier River from dam No. 20 (above Graham’s Ferry) to the crest of Bacon’s Falls. It was supposed that in a flood of 26 feet the irregularities due to the varied slope of the bottom shall disappear, and that the line drawn from the crest of the rapids above dam 21 to the crest of that below Graham’s Ferry and above the islands shall represent the slope of the river, and the section be that at this lower point. I find, by the Humphreys- Abbot formula, the discharge of the stream will be 65,652 cubic feet per second, say 66,000 cubic feet. I think this is a very safe estimate, for the abrupt bends between Graham’s Ferry and Bacon’s Falls and the islands opposite Rollinsburg mnst greatly impede the flow of water; accordingly the guard-walls and banks are recommended of such heights as to be above a discharge of this amount at the dams respectiv^ely. Mr. Hutton informs me that, in his arrangement of the slack- water system of the New River division, the locks can be used for the passage of boats during the passage of a Greenbrier flood. It is advisable that the Greenbrier may always be navigated when the New River can be, and accordingly the lock- walls, throughout their extent, are carried to the height of the abutment-walls and guard- banks in all cases. The plan of lock ad^ipted by Mr. Hutton for the New River has been taken as the model. The chamber- walls on the river side have been so calculated as to permit the lock to be emptied for repairs with a flood of 40,000 cubic feet per second in the river. a The mode of calculation is a modification of the empirical formula given by Lagrend Quoted from Mr. Talcott’s report of December 2:5, 1874. APPENDIX V. 711 for calculating the thickness of the chamber -wall of a lock, the inner and outer faces of which shall have the same batter, and is as follows : r is equal to the rise on the comb of the dam with a discharge of 40,000 cubic feet per second ;* h is the height from 1 foot below the bottom of the chamber-wall to the top of this rise ; /i' is the height of the top of the guard- wall to the top of the rise, and X is the total batter in the height h. The thickness of a wall of rectangular section which will just retain a column of water of the same height is iVo/b the specific gravity of the wall being supposed double that of water. A thickness of one-half h is then an excess of stability. The W'all shown in the figure will weigh per unit of length tt' {ah' all -j-x h), tt' being the weight of a unit of volume. The moment is / I./ I 7 1 TV {a h ah -j- xn ) — — The moment of a wall of rectangular section with height and a thickness of |/i, is Placing these two moments ennal and solvinsr with respect to x we have, + ^a^fih' -f- 4a'‘^ii^ -[- — 3a {zli -f- li) **= The base of the wall will then be equal to {a-\-2x.) The width, a, of the top of the wall has been taken at 5 feet. This break in the face of the wall will indicate for all time the danger-point for the rise in the stream, when the lock is empty and under- going repairs. The chamber-wall on the laud side is in all cases 5 feet on top, with a batter on front and on back of 5 on 1. At most of the dams a study of the character of the excavation will probably permit a decrease of this batter. f’he bottom of the wall is taken one foot below the bottom of the chamber. The value of h in the formula will then be equal to r -f- lift of lock -f- depth of water on miter-sill-}- J, and h' will equal the height of the guard-walls less the rise r. DETAIL DESCRirXION OF THE SLACK-WATER SYSTEM. Dam No. 4, “ 300 feet long and 23 feet high, and lock No. 4, of 13 feet lift, are located on a solid rock foundation, 1.32 miles below the month of Howard’s Creek. The lock is on the right, and connected with the hill-side by a guard-bank. On the left, a bluff below the railroad will form a natural abutment. In order to obtain a depth of 7 feet of water from the mouth of Howard’s Creek to this dam, a channel will be excavated through the shingle-bar just below the mouth of the creek. An estimate of the cost of this channel is included in that of the dam.” The amount of the excavation required for the lock is very large, (42,499 cubic yards.) It is probable that this amount may be materially decreased, if the location of the dam can be moved up stream. The bend in the stream a short distance below the lock makes a large excavation necessary to provide a proper exit from, and ap- proach to, the tail of the lock. This, and all other questions of location suggested in this report, can be readily determined during the progress of the work. The railroad at this point is out of danger. “About 200 feet below the dam the railroad crosses the river on an iron undergrade bridge of four spans. The bottom chord of the bridge is only 20 feet above the surface of the pool made by dam No. 5. One span of this bridge will have to be raised, in order to give sufiScient height for the chimneys of steamers, and made a through bridge.” Dam No. 5, 450 feet long and 20 feet high, and lock No. 5, of 9 feet lift, are located ^^on solid rock foundation, just above Clay’s mill-dam, and 3.01 miles below dam No. 4. The lock is on the right, and connected with the railroad by a guard-bank. On the left, an abutment of masonry and a guard-bank will also be required.” The railroad is only 8 feet above the comb of the dam, and an estimate is made for relocating it higher on the hill-side. Dam No. 6, 4l0 feet long and 21 feet high, and lock No. 6, 10 feet lift, “ are located on a ledge of sandstone 1.13 miles below dam No. 5. The lock is on the right, and con- nected with the railroad by a guard-bank. A rock bluff on the left forms a natural * Estimated by Francis’s formula, Q = 3.33 . I .r},\n which Q is the discharge in cubic feet ner second, in this instance is 40,000 feet ; I is the length of the dam. t If we make h' == O, the formula becomes ^ V a‘ -j- z — 3 a ~ 4 which is the empirical formula given by Lagrend, the column of water and the wall being of equal height. 712 REPORT OF THE CHIEF OF ENGINEERS. abutment. The railroad at this point is only 10 feet above the comb of dam, and will either have to be raised or protected for a distance of 0,000 feet by an embankment. A dike along the river would protect the low grounds and railroad.” Au estimate for raising the railroad is included in the estimate for the dam. An alternative to the construction of this dam suggests itself on an examination of the map, viz : To leave dam No. 5 by a guard-lock, and by means of a canal and 2 locks through the flats at Ronceverte, to reach the river below the mill at the elevation necessary to enter the pool of dam No. 7. An estimate of this alternate canal has been prepared for comparison, with the following results : Dam No. 5, first estimate $(12, 089 00 Lock No. 5, first estimate 11:1,935 68 Dam No. 6, first estimate 62, 668 40 Lock No. 6, first estimate 123,544 88 Total 362, 838 26 The total estimate for fhe alternate canal is $380,946.29, being an excess of $18,108.03 over the first estimate. The alternate estimate is but approximate, as the surveys do not supply complete data, for the alternate was not contemplated wlien the survey was made. I believe that I have allowed excessive quantities for this alternate line, and that an actual survey will show that the cost will be less than estimated. But even if the estimated cost should be not too great, the contingencies and damages would be less for the canal than for the dam and the lift-locks. The alternate estimates were for a canal 120 feet wide on the bottom, with side slopes 1 on 2, and the walls of the lower lock were supposed 20 feet above the comb of dam No. 7. The cutting for the upper canal was supposed to average 6 feet deep throughout, and that for the lower canal 4 feet deep. Of course the location of the lock is but con- jectural, and can only be made after survey. I recommend the alternate canal, subject to future survey, which should also provide more complete data for the change of rail- road necessary with dam No. 6. Dam No. 7, 350 feet long and 18 feet high, and lock No. 7, 7 feet lift, “are located on a solid rock foundation, 1.53 miles below dam 6. The lock is on the left bank, and will be connected with the high ground by a guard-bank; on the right a rock blulf under the railroad will give a natural abutment.” An estimate for the change of railroad is included in the estimate of the dam. I think it would be well, during the construction of the line, to examine this site with the view of locating the lock on the right bank, whereby the excavation for the lock might perhaps be much reduced. The bend in the river below is not advantageous for the lower approach as now arranged. Dam No. 8, 350 feet long and 18 feet high, and lock No. 8, of 9 feet lift, “are located on a sandstone ledge, covered with shingle on the right bank, but well defined on the left, and 1.32 miles below dam No. 7. The lock is on the right, and connected with the railroad by a guard-bauk. On the left an abutment of masonry and a short guard- bank will connect the dam with the hill-side. The railroad is here 17 feet above the comb of dam, and out of danger.” Dam No. 9, 320 feet long and 6 feet high, and lock No. 9, of 13 feet lift, “are located on a sandstone ledge, underlying red shale, 1.94 miles below dam No. 8. The lock is on the right, and connected with the high ground by a high bank. On the left a blutf of red shale will require a light masonry abutment.” This location is not the most advantageous for the approaches to the lock, and I would suggest further examination in this vicinity for an improved location. Dam No. 10, 435 feet long and 20 feet high, and lock No. 10, of 13 feet lift, “ are located on a limestone ledge, 2.14 miles below dam No. 9. The lock is on the left, and con- nected with the railroad by a guard-bauk. On the right a limestone clitf fqrms a nat- ural abutment. Between this dam and dam No. 9 the railroad crosses the river on an iron undergrade bridge, the bottom chord of which is only 14.5 feet above the surface of the pool, and the rail is only 34 feet above it. Oue span of this bridge will have to be raised and made a through bridge, which will give 30 feet clear above the surface of pool. Should more height be required, it will be necessary to change the grade in a tunnel which is not more than 400 feet from the abutment of bridge. The railroad at the dam is 15 feet above the comb, and out of danger.” Dam No. 11, 350 feet long and 22 feet high, and lock No. 11,11 feet lift, “ are located on a limestone ledge in Davis’s Falls, and 1.45 miles below dam No. 10. The lock is located on the right, and connected with the high ground by a guard-bauk.” Dam No. 12, 350 feet long and 24 feet high, and lock No. 12, of 13 feet lift, “ are located on a limestone ledge, 0.97 mile below dam No. 11. The lock is on the right, and con- nected with the high ground by a guard-bank. On the left a limestone bluff under the railroad forms a natural abutment. The railroad is hei'e 28 feet above comb of dam, and is out of danger.” The location of this dam is disadvantageous for the ap- APPENDIX V. 713 proach to the lock, and in general it may be said that such must always be the case at a bend, unless the bend be very slight. The boats, when the water is up, must enter the lock parallel to the thread of the current. The vicinity of this site should be further examined, or it may be that an alternate canal from dam No. 11 may render the construction of this dam (No. 12) unnecessary. An estimate of this alternate canal has beeu made. The comparison of the two sets of estimates gives the following results : Dam No. 11, first estimate $43,310 65 Lock No. 11, first estimate 125,819 18 Dam No. 12, first estimate 45,075 80 Lock No. 12, first estimate 152,456 98 Total, first estimate 366,662 61 Alternate estimate 337,449 64 Difference in favor of alternate 29, 212 97 This alternate is shown on sheet No. 1, and on the cross-sections of the canal-survey. I recommend this alternate, subject to an examination of the lock-sites and the char- acter of the stream in their vicinity. Dam No. 13, .320 feet long and 20 feet high, and lock No. 13, of 10 feet lift, “ are lo- cated on a solid limestone ledge, 1.56 miles below dam No. 12. The lock is on the right, and connected with the hill-side by a short embankment. On the left an abutment of masonry will connect the dam with the railroad, which is only 13 feet above comb of dam.” An estimate for the change of railroad is included. I would suggest a re-ex- amination of this site to decrease the great amount of excavation necessary for the lock. It may be that the location of the lock on the left bank, or the location of the dam a short distance down stream, may fulfill this object. Dam No. 14, 310 feet long and 26 feet high, and lock No. 14, of 12 feet lift, “ are lo- cated on a limestone ledge, 1.13 miles below dam No. 13. The lock is on the right, and connects immediately with the hill-side. On the left a masonry abutment will connect the dam with the railroad, which at this point is only 13 feet above the comb of dam.” An estimate for the change of the railroad is included in the estimate for the dam. “ Below this dam there is quite a number of large bowlders, which will have to be removed in order to make the channel of sufficient depth and width. This work is included in the estimate.” The excavation for this lock is quite great. I should anticipate that a location some 1,200 feet up stream would be an improvement in this particular, the lock being placed upon the left bank. Dam No. 15, 320 feet long and 20 feet high, and lock No. 15, of 10 feet lift, “ are lo- cated on a ledge of limestone, 1.04 miles below darn No. 14. The lock is on the right, and connected with the high ground by a short embankment. On the left an abnt- mentof masonry will connect the dam with the railroad.” An estimate for the change in the railroad is included in the estimate of the dam. The bend of the stream at this point makes it a disadvantageous location for the lock. I would suggest au examina- tion of a site about 1,000 feet up stream ; or it may be that a short canal may bo located from darn 14, which will render dam 15 unnecessary, especially if the location of dam 16 be changed as mentioned in the following description. Dam No. 16, 300 feet long and 22 feet high, and lock No. 16, of 9 feet lift, ‘‘are lo- cated on a smooth sandstone ledge just above Alderson depot, and 1.13 miles below dam No. 15. The lock is on the right, and requires a guard-bank 2,650 feet long to con- nect it with the high ground, and protect the bottom lands below the dam. On the left an abutment will connect the dam with the railroad, which is only 12 feet above the comb of dam, and ought to be raised to be out of danger. An estimate for the change of the railroad is included in the estimate for the dam. In order to reduce the height of the next dam below, a channel will have to be excavated through the shin- gle-bar just below the town of Alderson. An estimate for this work is included.” As an alternative to this dam, which will also avoid the construction of dam No. 17, the following suggests itsnlf, viz, to build a dam about 2,700 feet up stream, designated dam 16 Aon the map, (sheet No. 2,) then, leaving this dam by a guard-lock on the right bank, to locate a canal th’ongh the flats, and finally, as indicated on the map, to enter the pool of dam No. 18 below Muddy Creek. An estimate of this alternate has been made, with the following result: Dam No. 16, first esnmate $66,922 65 Lock No. 16, first estimate 131,211 IB Dam No. 17, first estimate 88,115 75 Lock No. 17, first estimate 126,380 28 Total, first estimate 412,629 86 Alternate estimate 444,086 6‘9 Difference against alternate 31,456 8^3' 714 REPORT OF THE CHIEF OF ENGINEERS. But the contingencies will be less in the alternate, the land-damages will be very much less, as well as prospective damages from floods, and the long guard-banks of dams 16 and 17 will be avoided. The alternate canal, as estimated for, is 120 feet wide on the bottom throughout. I recommend the adoption of the alternate line, subject to an examination for the site of dam 16 A and for the location of the locks on the island at Muddy Creek. Boring should also be made in the flats in order to assure the excavation of the canal in loose material. The following maybe found advantageous. To build dam 16 A somewhat higher than estimated, which would decrease the depth of cutting through the flat and perhaps render the canal from dam No. 14 expedient, thus avoiding the construction of dam 15. Da7n No. 17, 450 feet long and 23 feet high, and loek 17, 10 feet lift, “are located on a sandstone ledge below the mouth of Muddy Creek, 1.59 miles below dam No. 16. The lock is on the left, and requires a very long guard-bank or dike,” and that the rail- road be protected from floods. An estimate for the change of the railroad is included in the estimate for the dam. “On the right an abutment of masonry and a guard- bank connect the dam with the high ground. Below this dam the channel will have to be cleared of some shingle and bowlders in order to be safe. This work is included in the estimate.” Dam No. 18, 450 feet long and 20 feet high, and lock No. 18, of 11 feet lift, “ are lo- cated on solid rock foundation, 2.56 miles below dam No. 17. The lock is on the left, and will be connected with the darn above by a guard-baidi. The railroad must be raised for about 6,000 feet up from the dam.” An estimate for this change is included in the estimate for the dam. On the right a masonry abutment will connect the dam with the hill-side. Dam No. 19, 390 feet long and 19 feet high, and lock No. 19, of 9 feet lift, “ are located on a ledge of sandstone, 2.43 miles below dam No. 18, and about one mile above Haines’s Ferry. The lock is on the left, and is connected with the railroad by a guard-bank. The railroad must be raised for about 4,000 feet for protection. An estimate for this is included in the estimate for the dam.” Dam No. 20, 450 feet long and 18 feet high, and lock No. 20, of 8.5 lift, “ are located on a ledge of sandstone, 3.56 miles below dam No. 19. The lock is on the left, and con- nected with the high ground by a guard-bank.” It would be better, on account of the bend immediately below, if the location were changed a little up-stream. Dam No. 21, 420 feet long and 18 feet high, and lock No. 21, of 8.5 feet lift, “ are lo- cated on a sandstone ledge, one-fourth of a mile above Graham’s Ferry and 1.75 miles below dam No. 20. The lock is on the left, and connected with the high ground by a guard-bank. On the right an abutment of masonry and a short guard-bank connect the dam with the hill-side. Below this dam there are some bowlders and shingle which will have to be removed to make the channel of sufficient width and depth. About one-fourth of a mile below dam No. 21 the, railroad crosses the river on an iron through bridge of four spans, of 128 feet each. The bottom chord of the bridge is only 21 feet above the surface of the pool.” If the railroad interferes with the navigation it may be raised some distance each side of the bridge. Dam No. 22, 557 feet long and 17 feet high, and lock No. 22, of 8 feet lift, “ are located on a ledge of sandstone, at the head of Bacon’s Falls, or the Great Falls of the Green- brier, and 4.12 miles below dam No. 21. The lock is on the left, and connected with the bluff. On the right a perpendicular bluff of sandstone forms a natural abutment. At this point a canal has been located, which runs back of Bacon’s mill, and locks down by lock No. 23, of 11 feet lift, into the pool formed by dam No. 23. This canal is calcu- lated of sufficient width for two large boats to pass each other.” Dam No. 23, 400 feet long and 17 feet high, and lock 24, of 10 feet lift, “ are located on a ledge of sandstone at the head of the Little Falls of the Greenbrier, and 0.99 mile below dam No. 22. The lock is on the left, and connected with the high ground by a guard-bank. On the right a cliff of shale will have to be protected by a masonry abut- ment. Below this dam there are quite a number of bowlders of moderate size which will have to be removed.” Dam No. 24, 446 feet long and 22 feet high, and lock No. 25, of 11 feet lift, “ are lo- cated on a ledge of sandstone, one-fourrh mile above Carden’s White Sulphur Springs, and 0. 91 mile below dam No. 23. The lock is on the left, and is connected with the high ground by a guard-bank. At this dam, an island above, and one also below, will have to be excavated in order to make a channel of sufficient widrh.” An alternate line from above dam No. 22 is suggested, in detail, as follows: Build a dam, designated on the map (Sheet No. 3) dam 22 A, (foundation solid rock and length of dam 538 feet,) with a lift of 7 feet, locking into the pool of a dam to be built near where the crossing was made at Station 1876 4- 22. This dam would be about 22 feet high. In reply to a request for information, Mr. Talcott writes me as follows: “The rock ledge below Bacon’s Falls, where line crosses to left bank, is not lower than about 1,462, if as low, for the fall from foot of Bacon’s Falls to this point is very slight. A 30-foot dam could be built at this point and be very secure, for there are natural rock-abutments on both sides of the river. You would flood the present APPENDIX V. 715 site of Bacon’s mill, but as the water-power would be taken away, that would make little diffeience in the laud-damage.” Leaving this dam by a guard-lock on the left bank, a canal may be located through the flats below, until the line again enters the river below clam No. 25. The outline of these flats may be seen by referring to Sheet No. 4. It is seen (Sheet No. 3) that the line enters the river at 4 feet less elevation than the bottom of the lock at dam No. 24, thus permitting the lowering of No. 25, and decreasing the lift of lock 26 by that amount. An estimate of this alternate line has been made for comparison, with the following result : Dam No. 22, first estimate $43, 363 46 Lock No. 22, first estimate 108,729 18 Canal rouncl Bacon’s Falls 33, 312 40 Lock No. 23 105,793 18 Dam No. 23, first estimate - 46,994 60 Lock No. 24, first estimate 125,531 48 Dam No. 24, first estimate 71, 307 63 Lock No. 25, first estimate 115,561 28 Dam No. 25, first estimate 55, 992 48 Look No. 26, first estimate 120, 319 38 Total 826, 690 07 Alternate estimate 753,715 19 Difference in favor of alternate 62, 974 88 The estimate for this alternate is but approximate, as, the line not being contem- plated in the survey, our data are not complete. I have every confidence that the quantities are full and the estimate is sufficiently great. The canal around Bacon’s Falls is necessarily an awkward arrangement, from the little available space and the necessary lengths of the locks. Dam No. 23 is a disadvantageous location from the bend in the stream. Its location should be changed, or the excavation for lock 24 should be greater than estimated for. The laud-damage for dam No. 24 would be great, or for the guard-bank there should be substituted a dike 4,800 feet long, connect- ing this dam with the one above, in which case the estimate for dam No. 24 would be increased. I decidedly recommend the alternate line. This alternate being adopted. Dam No. 25 will be 426 feet long and 16 feet high, and lock at same (No. 26) will be of 7 feet lift. The site is on a ledge “of sandstone 0.96 mile below dam No. 24. The lock is on the left, and connected with the high ground by a guard-bank. On the right an abutment of masonry connects the dam with the hill-side.” Dam No. 26, 350 feet long and 23 feet high, and lock No. 27, of 11 feet lift, “ are located on a sandstone ledge 0.92 mile below dam 25. The lock is on the left, and is estimated for as connected with the hill-side by a guard-bank, though it might be advisable to connect it with the dam above by a dike along the river, which would increase the estimate for guard-bank.” This site is disadvantageous for the location of the lock, and if a dam be built near here I would suggest an examination for a site 1,000 to 1,400 teet below, when it might be advisable to increase the lift of the lock and decrease the height of dam No. 27 and the lift of lock 28. As an alternate to the construction of this dam the following is feasible, viz : Leave dam No. 25 by a guard- lock on the left bank, and then by a canal and locks to enter the river below dam No. 26, as indicated on the map, (Sheet No. 3.) This alternate will enter the river at 2 feet less elevation than the bottom of lock 27. An estimate for comparison gives the fol- lowing : Dam No. 25 $40,576 83 Lock No. 26 106, 198 88 Dam No. 26, first estimate 50, 403 55 Lock No. 27, first estimate 136,260 58 Dam No. 27, first estimate 42, 867 80 Lock No. 28, first estimate 123,083 28 Total 499, 390 92 Alternate estimate 455, 441 95 Difference in favor of alternate 43, 948 97 This estimate for the alternate is but approximate, but I am quite confident the quantities assumed are sufficiently full. I recommend the alternate line, subject to examination. If this alternate be adopted — Dam No. 27 will be 340 feet long and 20 feet high, and lock No. 28 will be 8 feet lift. The site is on a sandstone ledge 1.55 miles below dam No. 26. The lock is on the right, and connected with the high ground by a guard-bank. On the left a bluff of shale 716 REPORT OF THE CHIEF OF ENGINEERS. will need some masonry to protect it. A low dike may be necessary to decrease land- damages by protecting the low grounds above. (See Sheet No. 4.) Dam No. 28, 360 feet long and 22 feet high, and lock No. 29, of 10 feet lift, “ are located on a sandstone ledge 1.14 miles below dam 27. The lock is on the left, and connected with the hill-side by a guard-bank. On the right an abutment of masonry will con- nect the dam with the hill-side.” As an alternate to the construction of this dam, the following may be found avail- able, viz : Build a dam as an alternate to dam 27 (designated on Sheet No. 3 as dam 27 A) near station 2,152 of the transit-line. Then leave this dam by a guard-lock on the left bank, and by a canal and locks enter the river below dam No. 28. By this line the river may probably be entered at 6 feet less elevation than the bottom of lock No. 29, thus lowering the comb of dam 29. An estimate made for comparison gives the following: Dam No. 27, as recommended Lock No. 28, as recommended Dam No. 28, first estimate Lock No. 29, first estimate. ... Dam No. 29, first estimate. ... Lock No. 30, first estimate. ... $35,411 55 102, 135 88 53,923 45 116,921 38 42,400 75 112, 172 58 Total 462,965 59 Alternate estimate 486,624 82 Difference against alternate 23,659 23 The alternate estimate is only approximate, but 1 believe the quantities to be very full, and, there being one less dam to build, the contingencies will be less. I recom- mend this alternate line, subject to an examination. If this be adopted. Dam No. 29, 420 feet long, will be 15 feet high, and lock No. 30 will be 4 feet lift. The site is on a sandstone ledge, 1.5 miles below dam No. 28. The lock is on the right, and connected with the high ground by a guard-bank. On theleft'au abutment of masonry connects the dam with the steep hill-side. . Dam No. 30, 340 feet long and 22 feet high, and lock No. 31, of 11 feet lift, “are located on a sandstone ledge, 1.49 miles below dam No. 29. The lock is on the left, and connected with the hill-side by a short embankment. On the right an abutment ot masonry will connect the dam with the hill-side. Below this dam there are some very large bowlders which will have to be removed to give sufficient water-way. This work is included in the estimate.” The bend in the stream at this point renders this site a very disadvantageous one. I would suggest that the location be made lower down — the posirion designated dam No. 30 A, on sheet. No. 3. The notes as copied on the map indicate the probability that rock-foundation may be found anywhere in the vicinity. It may also be found that the elevation of the low'er level may be decreased, and that the lift of the locks Nos. 30 and 31 may, with advantage, be more nearly equalized by locating dam 29 farther down stream, perhaps about opposite station 54 of the line from the portal of the Great Bend tunnel. Dam No. 31, 326 feet long and 16 feet high, and lock No. 32, of 10 feet lift, “ are loeated on a ledge of sandstone, 2.67 miles below dam No. 30. The lock is on the right, and connected with the high ground by a short guard-bank. On the left a sand-bluff forms a natural abutment.” Dam No. 32, 465 feet long and 14 feet high, and lock No. 33, of 2.2 feet lift, “are located on a sandstone ledge, 1.51 miles below dam No. 31. The lock is on the right, and connected with the railroad by an abutment of masonry, and a guard-bank will connect the dam with the high ground.” This dam is the last on the Greenbrier River, and connects the slack-water system on that river with that of New River. At this point there exists a difference of recorded elevations between the notes of the surveys of the Greenbrier division and those of the New River. This difference is probably owing to a difference in the elevations assumed at the initial benches. The elevations in the Greenbrier survey are carried from the initial bench of the summit division at the mouth of Fork Run. This difference of recorded elevation is noted, Mr. Hutton tells me, in his report on the New River division. In the project of 1872 for the slack-water navigation of the Greenbrier River, it was proposed to tunnel the Great Bend. My opinion is opposed to this project. It must exceed in cost the slack-water round thd bend by a large amount. No estimate of it is submitted. The saving of distance does not necessarily produce a saving of time, for the rate of motion must be less within the tunnel than in the open river. The up-streanj approach would be difficult. Guard-locks would there be necessary, and the boats could not enter them across a current of even moderate velocity, and could not leave them in the ; APPENDIX V. 717 same circumstances without serious danger of accident from^ the leverage upon the boat from the down-stream pressure of the current. Then the accumulated lockage at the down-stream end of the tunnel (92 feet) would necessitate a flight of 8 or 10 locks at this point. The demands of the line would require that the flight be double. The cost of such a flight in the narrow creek-bed would be exceedingly great, and the ap- proach to the lowest locks must necessarily be difficult, as the objection to leaving or entering the locks across the current would apply similarly as at the upper portal. The total distance from the mouth of Howard’s Creek to dam No. 42 is 48.32 miles, and the total lockage 301.2 feet. A detailed estimate of the cost accompanies this re- port. The recapitulated estimate of the line as recommended is as follows : Dam No. 4 - $59,971 95 Lock No. 4 104, 730 98 Alternate canal at Ronceverte, to avoid — Dam No. 6 380, 946 29 Dam No. 7 67,437 50 Lock No. 7 114,523 48 Dam No. 8 41,753 40 Lock No. 8 126, 537 38 Dam No. 9 .54,743 70 Lock No. 9 157, 355 08 Dam No. 10 46,742 30 Lock No. 10 129,231 28 Alternate canal from dam No. 11, to avoid — Dam No. 12 337,449 64 Dam No. 13 41,7.30 80 Lock No. 13 195,360 98 Dam No. 14 80,946 00 Lock No. 14 156,240 58 Dam No. 15.' 53,849 80 Lock No. 15 141,454 08 Alternate canal, to avoid — Dams Nos. 16 and 17 444, 086 69 Dam No. 18 84,729 10 Lock No. 18 132,015 58 Dam No. 19 55, 634 75 Lock No. 19 129, 028 18 Dam No. 20 47, 235 00 Lock No. 20 106,118 38 Dam No. 21 45, 477 90 Lock No. 21 111,76108 Alternate, to avoid the construction of dam No. 22, locks Nos. 22 and 23, cana Iround Bacon’s Falls, dams Nos. 23 and 24, and locks Nos. 24 and 25. 753, 715 19 Alternate canal from dam No. 25, to avoid dam No. 26 455, 441 95 Alternate dam No. 27 and canal, to avoid dam No. 28 486,624 82 Dam No. 30 58,832 70 Lock No. 31 134,174 98 Dam No. 31 26, 104 40 Lock No. 32 148,226 88 Dam No. 32 29,270 40 Lock No. 33 •. 83,507 68 Total 5,682,990 93 Contingencies, 10 per cent 568, 299 09 6, 251, 290 02 Total estimate of the slack-water system recommended for this division, including 10 per cent, contingencies, is $6,251,290.02. Attention is invited to the remarks made in the report of the summit division, relative to the cost of look and dam masonry. The estimates for the darns have been made on the basis adopted by Mr. Hutton in the New River division. Tho classification of the excavation for the locks and dams is the same as adopted by Mr. Talcott in his report of 1874. The Chesapeake and Ohio Railroad maps were used to assist in the plotting of the railroad and of the bank of the stream not covered by our surveys. The estimates were made by Mr. J. L. Seager. 718 REPORT OF THE CHIEF OF ENGINEERS. Table showing length and height of dams, height of gnard-walls, and amount of water each dam will discharge before the flood will reach the top of the guard-wall. Number of dam. Length in feet. Height in feet. Height of guard- walls and hanks. Will discharge cubic feet per second — Remarks. 4 300 23 17 70, 000 5 450 20 14 78, 400 6 410 21 14.5 7.5, 300 7 350 18 *16 74, 500 8 330 18 17 75, 200 9 3-20 26 17 74, 600 10 435 20 14 7.5, 800 11 350 22 16 74, 500 12 320 24 17 74, 600 13 320 20 17 74, 600 14 310 26 17 72, 300 15 320 23 17 74, 600 16a 375 22 16 79, 900 Estimated height and length. 16 300 22 17 70, 000 17 450 23 14 78, 400 18 450 20 14 78, 400 19 390 19 15 75, 400 20 450 18 14 78, 400 21 420 18 14 73, 600 22a 530 15 13 82, 700 22 557 17 13 86, 900 23a 400 22 14 69, 700 Estimated length. 23 400 17 14 69, 700 24 446 22 14 77, 700 25 456 16 14 79, 500 26 350 23 16 74, 500 27 340 20 16 72, 400 27a 400 27 16 69, 700 Estimated length and height. 28 360 22 16 76, 700 29 420 15 14 77, 000 30 340 22 16 72, 400 31 326 16 16 69, 400 32 465 14 14 81, 100 GREENBRIER DIVISION — CANAL LINE. A survey was made for an independent canal simultaneously with the survey for a slack-water navigation. The following notes descriptive of the line of the canal-survey, taken by Mr. Talcott in the held, are copied from his note-book : “The line for the canal was begun at the point on Greenbrier River, above the mouth of Howard’s Creek, at which the northern tunnel-line would debouch should that be chosen. It crossed the river immediately, and can be crossed over either on aqueduct or slack-water, there being a good rock-ledge immediately at the point at which the line crossed. It then was run down on the right of the river, over the bottoms, until it crossed the James River and Kanawha turnpike, just below which point the hill closes into the river. Above this point the line was connected with the line from south- ern tunnel, it being thought best to run on the feeder-line and cross above the bridge, where a good rock-blutf forms a natural abutment, and a ledge of rock making on from the left bank would give a good foundation for piers.* “A connection was also made near the mouth of Howard’s Creek. From the poin t above named, or say station 109 of canal-line, down to station 180 the ground as a general thing is rough, steep, and formed of bowlders from the hill-side. The work here might be estimated as about one-half solid rock and one-half loose rock. At about the above-named station the line was crossed to the left bank and notes taken for either an aqueduct or slack-water, there being fine ledges for foundation. A crossing was also run below the railroad bridge, where it would be necessary to cross by aqueduct, as slack-water would raise the river so much as to endanger the railroad bridge. It * At the time the survey was begun I was under the impression that the canal-line should at once cross to the right bank of the Greenbrier Ri\er, and that the elevation of the summit would be 1,720 feet above the tide. This being assumed, I thought of crossing the Greenbrier on an aqueduct after tunneling |hrough the crest where the feeder-tunnel for the summit is located. These views are abandoned in the recom- mended location. — T. T. APi’ENDIX V. 719 will be necessary to cross the railroad at this point, as here it takes the right bank and occupies the only ground that could be used for canal-line. From the bridge down for about a mile there is pretty good ground, the bottom being narrow, but wide enough for canal. Thence down the river to a blulf about opposite Clay’s Mill the ground is steep and rough and of sandstone, either in ledge or bowlders. It would be safe tu take three-quarters solid rock and the rest loose rock in the cutting required here. “At the bluff opposite Clay’s Mill the best plan would be to tunnel for (say) 500 feet, as any other method would involve heavy cutting iu limestone or a high retaining- wall, which would be exposed to the strength of the current in times of high water. After passing the bluff the ground improves, and the bottom is wide enough for the canal for about 2 miles, when the bluffs close in again and render a crossing to tho right bank advisable. This crossing is at about station 495, and is located on a good ledge of rock ; but, owing to the bauk being low on both sides, if an aqueduct should be adopted, it might be well to run a little farther down on the left bank before cross- ing, so as to get a better height for the piers. “ From the crossing to the east portal of Second Creek Tunnel, Chesapeake and Ohio Railroad, the ground is open and a narrow bottom, very good for the canal. At this point the bluff comes iu again on the right, and for a half a mile the work would be very expensive if the line was kept on the river; so that, in order to save distance and probably some expense, atunnol-line was run across the neck, making about 2,800 feet of tunnel, and debouching on a narrow bottom on the river, saving about one mile and three quarters iu distance. A'l of the tunnel will be limestone, which iu the rail- road tunnel worked well and stands perfectly. After running a short distance in the bottom the bluffs close iu on the river again, and for three-fourths of a mile and more the work would be very expensive if kept out on the bluff. A sharp bend in the river below brings the bottom again on the line ; but I think that, on a full examination iu the office, a tunnel through this spur will be found cheaper and safer than a canal on the bluff, and save about three-fourths of a mile in distance. All of this bluff and si)nr are limestone. After running iu the bend above mentioned, a good bottom-land is struck, which lasts for about a mile, until ‘ Sinking Creek ’ is reached. Here the bluff comes down to the river again, and is very steep for about a quarter of a mile. A short tunnel may be found advisable here, or else a sort of gallery. The rock is lime- stone. “After passing this bluff the ground opens again and is pretty good for about a mile, when the bluff closes in again. Along here the line was run very high, in hopes of saving cutting by being above some of the bluffs, and alsoiu order to save work below where the flats are high. The rock here is still limestone. After passing these bluffs> the ground opens and is pretty good, though rolling for some distance. About station 900 the rock changes to sandstone, and is almost all in bowlders, though some ledges come down to the river, and limesfoue makes its appearance at times. Along here, iu mo't cases, the line was run too high, but the cross-sections will give the ground on which it will be best to put the canal, it being impossible to judge accurately what grade would be best until the line was run through. The ex- cavation,all the way down to station 1030. can be taken, as a large proportion is through bowlders, which, iu most cases, are so large as to be classed as solid rock, so that at least 75 per cent, of the excavation will be solid rock and the remainder loose rock and earth, the latter in small proportion. Ac station 1030, or about 1 mile above Al- dersou’s station, the bluffs recede from the river, and the bottoms are very favorable for canal. “At station 1063 a cro.ssing was made to the lefr. bank, so as to take the benefit of the bottom-lands on that side, where they are much more extensive than on the right, and a line was run and cross-secaon taken, the above station being equal to 0 of that line. The crossing was on solid rock and was for slack-water. Should it be thought best to cross on an aqueduct, it would be better to go about a mile and a half -farther down the river, where good rock-ledges can be found. There is a good ledge in the pool just below Aldersoii’s and above Hill’s store, on the right bauk, at which an aqueduct could be put. The slack-water crossing was made at the point named, as iu order to get across below it would have been necessary to build long dikes to protect the low grounds from overflow, whereas at this point a dike about 3,000 feet long will be all that is necessary to protect the right bank, and on the left thei’e is nothing to over- flow. The line was continued ou the right bank and all the necessary cross-sections taken, but the ground iu many places is of such a character as to render a canal very expensive. From Muddy Creek down for about 2^ miles, with few exceptions, the bluffs are close to the river. The rock, in most cases, is a shaly limestone, and would be easy of excavation. Just below the mouth of Griffith’s Creek the bluffs are highest and are of lamina' ed stone, which would staud very well in cuts, but I fear would not be good for tunneling. The nver is not wide enough to encroach on it much without danger, so that the canal could not be built out into it. “ Just at the point where Wolf Creek Mountain comes down to the river, on the left bank, tae ground opens ou the right, and is favorable for canal for about three-fourths 720 REPORT 01’ THE CHIEF OF ENGINEERS. of a mile, when it is rough and rolling again for about the same distance, and then becomes very steep. The rock is a bastard limestone and some pure limestone, not very hard to excavate. At about station 1420, the ground on right bank opens again, and becomes very favorable, being river-bottoms for about two miles. The line on the left was run down the railroad, and cross-sections taken all the way to the river. The ground to station 240 is very favorable, but at this point Wolf Creek Mountain comes down to the river, and for about 3,000 feet the ground is rough and precipitous in some places. At this point the railroad runs close to the river, and the country road just above it, so that there is no room for a canal without throwing the railroad into tunnel, and, it may be, tunneling for canal for a short distance. Cross-sections have been taken of this point to show accurately what will be required, but if the line on the left bank is adopted, the best plan will be to pass this point by slack-water, and for that purpose a dam was located at the upper end of Riff’s low grounds, and the necessary notes taken to determine the cost of such a work. From this point down to station 335 the line is cross-sectioned, and at that place a dam is located to cross the i'ofiles, &c., refer to elevations above tide- water at Richmond, Va. The basis for all the levels was a bench-mark on the left bank of the Greenbrier, near our initial jioint, established by the party under the late Mr. Ed. Lorraine, in 1872. This bench-mark was, in turn, established by levels based upon a bench-mark established by parties under Mr. W. R. Hutton, in 1870, on the left bank of the Green- brier, near the mouth of Howard’s Creek, the levels having been brought over the Alleghanies hy these parties. During the summer of 1874 other lines of levels were brought over the Alleghanies by parties under Lieut. Thomas Turtle, United States Engineers, which were found to difler somewhat from those of 1870, and upon their prolongation to the mouth of the Greenbrier this difference was increased largely, so that, according to the last levels run, our bench-mark at initial point was 4.8 feet too low. This discrepancy was not discovered until the field-work and notes of both divisions were completed, and, in order that the drawings might correctly represent. the contents of the note-books, the original levels have been retained and marked, so tliat the water-surface at dam No. 32 of the Greenbrier division appears to be 4.8 feet higher on the maps and profiles of that division than the level of the same point ap))ears to be on the maps and profiles of the New River division. In other words, to reduce the levels of New River divis- ion to correspond to those of the Greenbrier division, 4.8 feet must be added to them. GENERAL FEATURES OF THE COUNTRY. The New River, throughout the yiortiou which was embraced within the limits of Ihis survey and report, presents in various degrees of development in its different sec- tions the contracted valley and sudden changes of slope usually found in streams flow- ing through mountainous districts. Between the Greenbrier and Kanawha Falls, the river may be divided into four sections, possessing, in varying and increasing propor- tions, from the upper to the lower end, the characteristics referred to. The fiist section (about 15 miles long) extends to Meadow Creek, with an average fall (excluding the abrupt fall at Richmond’s) of about 7 feet per mile. The valley, thongli narrow, presents usually on one side or the other strips, at least, of bottom- land, and the hill-slopes are not continuously precipitous. The second section (about 25 miles long) extends to Sewell, with an average fall of about y feet per mile. The valley becomes more contracted; bottom-land (properly APPENDIX y. 729 so called) lias almost entiiely disappeared, its only ropresentatkve being short stretches of high bench-land. Vertical cl ids impinge more freciueutly upon the water’s edge, and the hill-slopes are everywhere ragged with masses of rocks detached from the cliffs, always found at short distances in the rear. At this })oint (Sewell) is usually said to commence the “gorge” Avhere the river breaks through the high lands couuectiug the Gauley Mountains with those of Coal River. The third section (12 miles in length) extends to Hawk’s Nest, with an average fall of 17 feet per mile. Throughout this, as well as the following section, the river has cut its way over a thousand feet deep through the highlands, and were it not for the more friable nature of the materials encountered, (sandstones and shales,) Avould pre- sent all the features of a veritable canon as found in the metamorphic regions of the Great Basin. Both valley and river-bed are very much contracted, with steep slopes of loose rock thrown down from the cliffs. The water-way is frequently clogged with large masses of rock, through which the river alternately rushes with violence in many narrow streams, or eddies in deep and sluggish pools. The fourth section, extending to Narrow Falls, about seven miles in length, has an average fall of 19 feet per mile, and is the culmination of the gradual increase of rug- gedness, contraction, and slope. Though the average fall, as stated, is 19 feet, there are several miles in which the fall is nearly 80; there can hardly be said to be any valley ; the vertical and at times overhanging cliffs rise abrubtly from the w^ater’s edge, or are at best oulj^ separated from it by a sloping mass of loose rock of all shajies and sizes; the masses of rock cumbering the water-way appear at times almost to have absorbed the river, which truly offers at first sight small chance for improvement. From this point (Narrow Falls) to the Kanawha Falls the river extends in a wide pool of varying depth, the side-hills come closely down to the river until the Gauley enters on the right bank, from which to the falls the valley expands on that bank in a wide fiat. The width and depths of water-way on this whole division vary so frequently and so greatly, that it would be impossible to give an accurate general statement of them ; it may, however, be roughly said in the first section the width is from 400 to 80t) feet, with a depth of from 7 to 12 feet in the pools; in the second section, from 250 feet to 400 feet, with the same depths in the pools; and in the third and fourth sec- tions it rarely ever exceeds 250 feet in width, and is often less than 200 feet, with depths in the pools as great at places as 30 to 40 feet. It flows in alternating pools and rapids, with depths in the latter varying from a few inches in the ujiper sections of the river to 10 feet in the lower portions. The only vertical falls encountered are Richmond’s, about ten miles below the Greenbrier, where the river falls 28 feet, 15 feet being nearly vertical over a ledge of hard, conglom- erate sandstone; and Kanawha Falls, where it descends 194- feet over a ledge of sand- stone. In both cases the river has a width of over 1,000 feet on the high-water Crest- line, and immediately above the river is very shallow. At Richmond’s Falls, the hill- slopes, or cliffs, come closely down to the water’s edge on both sides of the river; at Kanawha Falls, the hills come down closely on the left bank, and a wide, low bottom extends along the right bank. The Chesapeake and Ohio Railroad occupies the right bank of the river, from the Greenbrier to Hawk’s Nest, and the left bank thence to Kanawha Falls. The rock ex- posed throughout is mainly a compact sandstone, iu nearly horizontal beds, occasion- ally overlaid with real shale, and sometimes merging into a hard conglomerate, as at Richmond’s Falls. The hill-slopes are generally formed of debns from the a Ijaceut cliffs iu the upper portion, to some extent mixed with alluvial deposits; but lower down they form, as before stated, almost pure masses of detached rock. A large majority of the shoals are composed of loose rock and bowlders forming dams, over and through which the river flows, and which have apparently (after the manner of the formation of “ pot-holes ”) abraded and broken up the bed-rock to great depths, rendering these apparently natural sites for dams actually the least desirable, in so far as foundations are concerned. For this reason it will be found that in the present project the dams are frequently located in the deep water, above the shoals, giving an increased height of masonry, but securing undoubted rock-foundations. Except at the shoals, the soundings indicated generally a bottom of rock iu ])lace, with little or no deposit over- lying it. The bottom-lands have generally a loose, sandy soil, in most cases inti- mately mixed with debris from the cliffs. The side-hills are almost entirely destitute of anything like soil in masses, such as there is being swallowed up iu the crevices of the loose rock which everywhere cov- ers their slopes. Nevertheless, walnut and poplar timber, of large size and excellent quality, abounds on these slopes throughout the whole d stance. The sandstone everywhere found is of excellent quality for building purposes, and easily quarried, in immediate proximity to the sites where required for use. The question of water-supply does not enter into the discussioti of any method of navigation-improvement, as its abundance is undoubted, its superabundance during freshets being the most formidable obstacle. 730 REPORT OF THE CHIEF OF ENGINEERS. Until the construction of the Chesapeake and Ohio Railroad, owing to the scarcity of arable land, the valley between the Greenbrier and Kanawha Falls was an unknown land, except to parties of surveyors or hunters who at long intervals forced their way down it. Owing to this absence of settlement, we have not only no accurate record as to the volume, frequency, and duration of floods, but not even the ordinary hearsay information usual in such cases. The wild and gloomy appearance of the river through the “ gorge,” as seen from the cliffs above, has doubtless caused an exaggeration of their heights and violence, great as they undoubtedly are ; but that they attained great heights, with extreme rapidity, is all that we really do know. During the sum- mer of 1861 occurred a freshet which is admitted by the few old settlers scattered throughout the valley to have been the greatest known for forty years ; and of this some few records of heights are obtainable, though that they are uucertaiu may be admitted, when it is known that at one point a difference was found of 7 feet in the elevations, as given by different persons claiming to have seen and noted the flood at its height. However, this flood, and the measures of it thus obtained, have, since its occurrence, been used as the standard of reference by the Chesapeake and Ohio Rail- road Company and for all projected improvements of the river, including the projects herewith submitted. To determine its volume, reference has been had to its reported observed height of 8 feet just above the crest of Richmond’s Falls. It has been assumed that for a length of 1,200 feet (which is less than the develop- ment of the crest-line by probably 15 per cent.) the discharge was equivalent to that over a weir having 8 feet depth above the comb. This would give a volume for this flood of about 90,000 cubic feet per second, which is the amount that has been assumed heretofore as well as in this project to determine proportions of guard-walls and banks and the heights of dams with reference to the railroad. From observations at other points of the water-marks of this flood, combined with the cross-sections of the valley, I am inclined to the belief that this estimate of quantity is exaggerated. It is true that the data available are too rude for other than a very approximate solution of the question, and altogether insufficient to justify the construction of works on any other basis previous to the procurement of more reliable information. The matter is referred to simplj" to call the attention to thepossiUUty of greatly lessening the cost of any scheme of improvement by the aid of more full and reliable data as to the floods. The reasons for a belief in this possibility are twofold ; in the first place, the observed height of 8 feet at Richmond’s Falls was toward (if not on) the left bank, from which extends, for more than half the total width of the river an irregular island, (or peninsula,) im- mediately below the crest of the falls, at some points really connected with it, at many nearly of equal height, and covered throughout with loose rock and a dense growth of timber, which during this great freshet was in full foliage. It seems almost impos- sible that this obstruction to the free escape of the water should not have caused an elevation of the level of the crest on this part of the line over that portion which afforded a free discharge, and consequently the assumption that 8 feet was the average depth is an overestimate. Secondly, a comparison of the observed heights of the flood, at some points, with the cross-sections of the valley, with liberal allowances for the unknown and for errors in the known quantities, indicates a volume not ex- ceeding 60,000 or 65,000 cubic feet per second. Should this assumption proveto be true, it will result that the project as now presented will tax the work with difficulties of execution and operation in excess of any probable requirements. The evidence of flood-heights in the “ gorge” of 50 to 60 feet was, at the time of this survey, confined to a single log lodged in the cliffs, in the right bank, below Cotton Hill station. This undoubtedly presented evidence of abrasion, as if by contact with water and rocks, yet it may have come down the hill and not up, for almost immediately below it (about 30 feet) is a cave or hole in the cliffs, so situated that it would seem impossible for all, if any, of the drift entering it to escape ; yet a careful investigation failed to discover any signs of such deposit within it. SLACK-WATER IMPROA'EMENT. The general arrangements for this method of improvement, as set forth in the report of survey in 1872, made under your direction by the late Mr. E. Lorraine, necessarily formed the basis of operations for this re-examination. There remained only to be de- termined the feasibility and costs of eliminating from that project certain features (mainly relating to the foundations for and length of dams) which were considered objectionable. The question of water-supply, as before stated, does not enter into the question, ex- cept as to protection against its superabundance. The result of the re-examination will probably be more readily appreciated by a summary of the leading features of the former project and that now presented for con- sideration. As arranged and estimated for in 1872, the plan embraced (between the mouth of Greenbrier and the pool below Kanawha Falls) 38 dams, 9 of which were located on foundations of loose rock. Their average length on crest was 581 feet; mini- APPENDIX V. 731 mum length, 317 feet ; and average height above foundations, feet. (The maximum lift at any one lock being 25 feet.) The project, as now presented, proposes 33 dams, all located on solid-rock foundations, with an average length of 709 feet, a minimum length of 550 feet, a mean height of 24 feet, and a maximum lift of 22 feet, making, as compared with the former arrangement, a decrease of 5 in the number of dams, or 13 per cent. ; an increase of 128 feet in mean length, or 22 per cent. ; an increase of 233 feet, or over 60 per cent., in minimum length ; and a decrease of 3 feet in maximum lift, without material change in the average height above foundations of the dams. The diminution of the number of dams, without increase of height, is proposed to be obtained by a resort to sections of lateral canal around some of the greater descents, and by the exca-vation of channels through the shoals next below the dams, in order to utilize the natural depths of water found at their feet. The increase of mean and minimum length is proposed to be obtained by building the dams obliquely to the thread of the current in one or more branches, and by the location of the locks and lateral canals as far into the hill-sides and away from the low-water border as circumstances will permit. This latter arrangement affords the greatest possible development of crest-line, as it obtains nearly the entire distance between the level contours on either side of the valley corresponding to the reference of the comb of the dam. It evidently, however, requires very considerable excava- tions to form approaches to the locks, and in some cases the removal of projecting points of the side hills above and below the dams, to permit the free access and dis- charge of flood-waters. On the other hand, in addition to the increased length of the dams, it partially removes the locks from the more violent assaults of the floods, and is believed to afford the nearest practical approximation to compliance with the sug- gestion that the locks should be located in lateral ravines, entirely removed from the river-bed, for the reason that these are always narrow, rugged, and precipitous, liable to frequent and violent floods of short duration, which bring down masses of stone and timber, rendering them, if anything, more dangerous than the main stream to the stability and security of the works. Examinations for tunnels were made across all the principal bends of the river that would be accessible to either a canal or slack-water improvement. The location of the Chesapeake and Ohio Railroad would prevent (with any reasonable degree of economy) the approach to any possible tunnel occupying the same side of the river as the road, and consequently no examinations were made on that bank, except at the Stretcher’s Neck Tunnel. Tunnel-line No. 1 passes through the first bend above Stretcher’s Neck, leaving the river above Quinimont Station and striking it again immediately opposite the entrance to Stretcher’s Neck tunnel. The distance by river is 11, 600 feet And by tunnel 3, 600 feet Or, tunnel-line saves 8, 000 feet The tunnel-line will cost about $1,750, 000 00 And the river-line will cost about 108, 000 00 Or, cost of tunnel-line will exceed line around by 1,642,000 00 Tunnel No. 2 is through Stretcher’s Neck bend. The distance by river is 19, 000 feet And by tunnel 2, 000 feet Or, tunnel-line saves 17, 000 feet The tunnel will cost about.. $1,0.37,000 00 And by river will cost about 570, 000 00 Or, tunnel-line exceeds in cost by 467, 000 00 Tunnel No. 3 is about 5 miles below Stretcher’s Neck, and cuts off Buffalo Shoals. The distance by river is 14, 000 feet And by tunnel is 4, 950 feet Or, tunnel-line saves 9, 050 feet 732 REPORT OF THE CHIEF OF ENGINEERS. The cost by tunnel is about $1, 230, 000 And by river is about 237,000 Or, tunnel-line costs in excess 993, 000 Tunnel No. 4 is through the first bend below Hawk’s Nest, and is the first one located on the right bank of the river. The distance by river is 12, 000 feet And by tunnel-line. 6, 400 feet Or, the tunnel -line saves 5, 600 feet The cost of tunnel-line is about $1, 529, 000 And the river-line about 453, 800 Or, tunnel cost is iu excess by .1, 075, 200 Tunnel No. 5 is also on the right bank, and cuts off the Blue Hole Bend. The line would, at lower end, strike the river at the foot of Narrow Falls, above darn No. 32. The distance by river is 12,500 feet And by tunnel is 7, 760 feet Or, the tunnel-line saves about 4, 740 feet The cost of tunnel-line is |2, 087, 500 And the cost by river is 427, 178 Or, the tunnel-line exceeds in cost about 1, 660, 322 Tunnel-lines Nos. 2 and 4 will enable one dam at each place to be dispensed with. The others save nothing in this respect. All the tunnels will require a guard-lock in addition to the same number of lift-locks as the lines by way of the river. If the adoption of these tunnel-lines would serve in any conuderable degree to raise the standard of utility of the slack- water schema, or, iu other words, if their use would enable the whole division to be navigated during floods of gi’eater volume than could be done without tham, their construction might be advisable even at a greatty in- creased first cost. But this would manifestly not be the case, for the reason that those portions of the route not aftected by any of the tunnels possess all the characteristics of the portions avoided by them. The river above tunnel No. 1 presents no greater facilities for the operation of a slack navigation than is fouiifl iu the bend cut off bj^ the tunnel, and the same is true of the portions between Nos. 2 and 3 and Nos. 4 and 5. Those portions of the river avoided by lines Nos. 4 and 5 undoubtedly offer very formidable obstacles to the con- struction and operation of any scheme of navigation improvement, and here, if any- where, the resort to tunneling might be true economy. The ihver at these places has, for several miles, an average fall of 26 feet per mile, one of the miles, avoided by tun- ]iel-line No. 5, having a fall of about 30 feet. Yet even here the intermediate portions not aflected by the tunnels are equally unfavorable with those avoided, and for the gorge,” as elsewhere, the only real gain possible by means of tunnels isthat of dis- tance or time. Assuming 20 minutes as the time required to pass through one of these large locks, and the average rate of travel by boats iu the pools as 3 miles an hour, tlie extra guard-lock at each tunnel will consume an amount of time e(i[ual to 1 mile of d.stance . If we further assume, as is reasonable, that the rate of travel through the tunnels and approaches will be only 1^ miles per hour, there will result the folio wm^ comparison of times by the several tunnels and by the river, ex«5ladiag lockages common to both routes: Minutes. Through tunnel No. 1 By river Through tunnel No. 2 By river Through tunnel No. 3 By river Through tunnel No. 4 By river Through tunnel No^5 By river 47 44 35 72 57 53 51 37 60 37 APPENDIX V. 733 So thatj even allowing a much less time for lockage and a greater rate of speed through tunnels and approaches than above, the fact remains that there would be no gain in time by the use of any of the tunnels except No. 2, (through Stretcher’s Neck.) For these reasons, though estimates of cost are submitted for tunnels, only that through Stretcher’s Neck is eqabrased in the main estimate showing the total cost of the slack- water scheme. The ultimate utility of the whole project is evidently governed by that of its worst portion, which is the gorge between Sewell and Narrow Falls, and the utility of this section evidently depends upon the number of days during which navigation would be suspended by reason of floods rendering the pools impassable. The data necessary to a very positive determination of this question are not to bo had; yet sufficient is known to enable an approximate solution to be made, which will be satisfactory just in proportion to the certainty that the assumptions are all made against, rather than in favor of, the slack-water navigation. Under certain assump- tions, which are in harmony with the facts, in so far as known, a table (hereto ap- pended) has been prepared, showing the discharges and velocities corresponding to every foot in depth, from 1 foot to 14 feet, on the comb of dams similar to those pro- posed for this part of the work. Assuming that a velocity of four miles an hour at the upper ends of the pools will represent the limit against which such boats as would probably be used on this improvement could be propelled, it will be found by reference to the table that, for the duration of any flood discharge exceeding 37,000 cubic feet per second, navigation by slack-water would be entirely suspended throughout the gorge. The only definite information on record as to the height and duration of floods is confined to observations made by Mr. E. M. Tutwiler, civd engineer, (at the time in the service of the Chesapeake and Ohio Eailroad Company,) who has kindly furnished a profile of his results, which is appended l|ereto. These observations were made near Buffalo Shoals, about 5 miles below Stretcher’s Neck, and extend over a period of about nine months, from June 12, 1872, to March 17, 1873. The season appears to have been prolific in frOishets, no less than ten being noted exceeding 6 feet in height above low water. The extreme range was 19 feet above this plane, and for 247 days out of 278 days, or about six-sevenths of the time, the river was at various stages above low water. If these observations had included measurements of slope or velocity, or had similar ones been made above and below, so that the volume of discharge at the various heights could have been determined, the question would have at once been settled for all similar years. As this was not done, we cark only approximate a solution, based upon assumptions always certainly against the slack-water. The flood of 1861 (according to marks given by persons who resided in the vicinity at that time) attained at this point a height of about 22 fbet, and according to the es- timates hitherto used, based upon its depth on the crest of Richmond’s Falls, its vol- ume was about 90,000 cubic feet per second. If we assume that at one-half the height the volume of discharge would be , (which is certainly a liberal allowance,) it will result that during the continuance of floods exceeding 11 feet in height at this place the navigation by slack- water 'would be suspended throughout the gorge. An examination of the records of Mr. Tufcwiler’s observations shows that during the nine months covered by them there were four days in January and five and one-half days in February, or nine and one-half days in all, during which this suspension of navigation would have obtained in 1872 and 1873. Whether this season was a fair average as to high waters we cannot tell with cer- tainty, owing to the absence of records, but it certainly could not be considered a low-v^ater season. MECHANICAL STRUCTURES. Locks . — The drawings herewith subraftted show in detail the type of locks, gates, valves, &c., proposed to be used as applied to the mean or average lift of 15 feet. In accordance with the resolution of adopting for the locks of the canal division a width of 24 feet, the plans for the slack-water locks have been made 250 feet by 48 feet in the chamber, instead of 240 feet by 40 feet as hitherto proposed and estimated for. The details of gates, valves, &c., are of course only intended as studies, and it is hardly to be doubted but that in the course of construction many improvements and econo- mies can and will be made. The locks throughout are proposed to be protected against submergence by iloods of 100,000 cubic feet per second, or 10 per cent, greater than that of 1861. The walls and gates are proposed to bo arranged to permit the use of the locks when the river is at the maximum navigable height in the pools. Owing to the fact that the locks are located as far as possible from the water’s edge, guard- banks are short and unimportant, the wing-walls frequently being sufficient. It is proposed to found all locks on solid rock, and in preparing the estimates its position was determined, where concealed from view, by reference to its exposure in the river- 734 REPORT OF THE CHIEF OF ENGINEERS. bed and in the adjacent hill-sides. The drawings show the upper gates resting on breast- walls ; this is proposed to be done only when the excavation is in solid rock, or in other special cases; generally, the upper and lower gates will have the same height, in order to prevent a water-fall at the head of the chamber when using the valves in the gates to fill the lock. The lock-flooring of timber and concrete, shown and estimated for in each lock, is only proposed to be used where the rock is seamed or imperfect, a matter only to be determined by actual experience. It is proposed to build the chamber-walls of rubble-masonry, pointed off” on the interior, and the head and tail walls of random-ranged masoury, dressed on exposed faces. The walls have been designed to resist the pressure of the water when at flood- height, supposing the chamber filled with water to the lower level. The European practice appears to be to consider the lock-chamber empty and the exterior water at flood-height; but as this method would greatly increase the volume and cost of masonry, and could only be demanded by the possible necessity of empty- ing and repairing the lock-chambers during great floods, it has been thought an unne- cessary expense for this project, as the emergency, should it arise, could be met by bracing the walls from the interior. The details of the wrought-iron gates are adapted to this project, mainly from the drawings of a lock-gate at Clarendon, (France,) as illustrated by De Lagrene, in his “ Cours de Navigation Interieure.” It is expected to fill and empty the locks by means of iron pipes in the side-walls, passing behind the hollow quoins, as well as by valves in the gates. It is estimated to fill a 15-foot lift-lock by the pipes alone in 12 minutes. A method of maneuvering large lock-gates with certainty and celerity appears to be, as yet, undiscovered. The arrangements commonly in use are beams, or spars, for push- ing shut or pulling open the gates, chains working over drums on the side-wa Is, twn chains being required for each leaf, a toothed iron arc secured to upper part of gate at about one-quarter the width of the leaf from the “ heel-post,” and working into a cor- responding arc imbedded in the side-wall, the arrangement being worked by crank and pinions ; and finally, by a system of gearing worked from the top of the gate, actuating a wheel or roller traveling on a track on the floor of the gate-recess. A necessary condition to the successful working of any of these plans requires that the roller-track near the outer end of the gate (which is common to all large lock-gate plans as a point of support, &c.) should remain unobstructed by deposits of sediment or debris of any kind. If this condition is maintained it would seem that the system of moving the gate by machinery attached to the outer end, and worked from the top of the gate itself, ought to be the more compact and etfective mechanical arrangement for effecting the desired object. The drawings show this method of working, as well as by a simple roller to be moved by either chains or booms. The booms are cumbersome and greatly in the way in a narrow valley like that of New River. The chains interfere with close joints of gate and miter sill, and are liable at any time to be effectually blocked by small chips or stones. The short iron arc requires immense power to compensate for its want of leverage, and a long one would break of its own weight. New River is not a sediment-bearing stream, and in it, if anywhere, it may be ex- pected that a track on the lock-floor will remain unobstructed. The toothed arc on the floor in this project is, for greater security, proposed to be raised above the floor on cast-iron brackets. Dams . — The dams are jy’oposed to be built of rubble-masonfy, pointed off on exposed faces with a heavy stone coping, covered with timber. The estimates contemplate giving them a section of not less width than 10 feet, top or bottom, with a general thickness at bottom of about -fn- the height. The abutments are proposed to be of same character of work and material as the dams. The estimates also include for each dam the excavation of a trench 2 feet deep over the whole area of the foundation, and its filling with concrete; this, of course, will only be needed where the rock is seamed or imperfect. Five of the dams are located on rock of a slaty structure ; the remainder on compact sandstone. This division comprises 6 guard-locks and .51 lift-locks, of an average lift of 14^ feet, and 33 dams having a mean length of 709^ feet and a minimum length of 550 feet. SYNOPSIS OF PROJECT AS NOW PRESENTED. The locks are proposed to be made 250 feet by 48 feet in the chambers ; canal prisms, where possible, 102 feet wide at water-line, and never less than 60 feet for very short distances. Tunnels are proposed to be made 54 feet wide at water-surface, 9 feet deep below it, and 33 feet high above it. APPENDIX V. 735 Lock walls and gates are proposed to be arranged to permit navigation with 7^ feet on a dam, 600 feet long, and are arranged to exclude floods of 100,000 cubic feet per second. Commencing at lock in dam No. 32 of Greenbrier division, on the right bank of the Greenbrier, (to avoid flooding the Hinton bottoms, as would have been done by dam No. 1 of the survey of 1872,) a canal is traced 6,850 feet through the Hinton bottoms to lock No. 1, of 12 feet lift; thence the canal is continued 6,275 feet, partly in the river and partly over low points of bottom-land, to station 138, below the town of Hinton, where the line enters a pool 2,700 feet above dam No. 1. Dam No. 1, 14 feet high and 1,010 feet long, is just above Tug Shoals; foundation on sandstone. Lock No 3, which is a guard-lock, is on left bank, and gives entrance to a canal around the shoals, 1,900 feet long, to locks Nos. 4 and 5, of 10 and 13 feet life respectively, which lock into the pool from dam No. 2. Dam No. 2, 11 feet high and 1,000 feet long, is at head of Brooks’s Falls, 9,500 feet below lock No. 5; foundation, sandstone. Lock No. 6, of 11 feet lift, is located on the lefn bank, and locks into the pool formed by dam No. 3. Dam No. 3, 16 feet high and 960 feet long, in two branches, is at Bragg’s Island, 6,300 feet below lock No. 6; foundation on sandstone. Lock No. 7, of 9 feet lift, is on the right bank, and locks into pool formed by dam No. 4. Dam No. 4, 12 feet high and 1,290 feet long, is located at Richmond’s Falls, 12,100 feet below lock No. 7 ; tlie foundation is on hard conglomerate, with slaty lamina, to avoid the long section of canal on the left bank, as proposed at this place by Mr. Lor- raine. It is now proposed to locate the locks (with intermediate basins) close to the right bank, cutting directly through the crest of the falls, and utilizing the deep pool above the debouche of Mr. Lorraine’s canal. Lock No. 8, of 13 feet lift, is on the right bank at the dam. Lock No. 9, of 14 feet lift, is 300 feet below No. 8, and connected with it by a canal of that length. Lock No. 10, of 13 feet lift, locks into the pool formed by dam No. 5 ; it is connected with No. 9 by a basin .350 feet long. Dam No. 5, 16 feet high and 770 feet long, is located 12,000 feet below lock No. 10 ; foundation, sandstone. Lock No. 11, of 12 feet lift, is on the left bank, and locks into X)ool formed by dam No. 6. Dam No. 6, 15 feet high and 760 feet long, is located at head of Meadow Creek Bot- tom, 7,700 feet below lock No. 11; foundation, sandstone. Lock No. 12, on the right bank, is a guard-lock to enter a canal 2,100 feet long through the Meadow Creek Bot- tom. Lock No. 13, of 10 feet lift, is at lower end of canal, and locks into the pool formed by dam No. 7. Dam No. 7, 16 feet high and 600 feet long, is located 6,200 feet below lock No, 13 ; foundation of slaty structure overlaying sandstone. Lock No. 14, of 15 feet lift, is on the left bank ; connects by a cut 600 feet long with the pool formed by dam No. 8. Dayn No. 8, 17 feet high and 1,640 feet long, is 10,100 feet below lock No. 14; founda- tion, sandstone. Lock No. 15, of 15 feet life, is on the left bank, and locks into pool formed by dam No. 9. Dam No. 9, 20 feet high and 820 feet long, is located at and above mouth of Glade Creek, 12,300 feet below lock No. 15 ; foundation in firm slate. Lock No. 16, of 12 feet lift, is on right bank, and connects with pool from dam No. 10. Dam No. 10, 20 feet high and 600 feet long, is 7,800 feet below lock No. 16; founda- tion of sandstone and shale intermixed. Lock No. 17, of 13 feet lift, is on the rigUt bank, and connects with pool from dam No. 11. Datyi No. 11, 22 feet high and 640 feet long, is 8,100 feet below lock No. 17, at the head of low bottom above Quinimout; foundation on hard shale. Lock No. 18, of 10 feet lift, is on right bank, and connects with a canal through low ground, 1,100 feet long, to lock No. 19 of 10 feet lift, which connects with pool from dam No. 12, giviug entrance to Stretcher’s Neck Tunnel, 11,400 feet below dam No. 12. Dam No. 12, 27 feet high and 630 feet long, is located below Stretcher’s Neck Tunnel, on a sandstone foundation. Its function being simply to back the water up to tunnel- entrance, no lock is provided for it ; foundation, sandstone. Lock No. 20 is a guard- lock, located on the right bank, opening into approach-basin to Stretcher’s Neck. Stretcher’s Neck Tunnel is 1,320 feet between portals, and is followed by lock No. 21, of 19 feet lift, locking into a basin 310 feet long. Lock No, 22, of 18 feet lift, is at lower end of basin, and is immediately followed by lock No. 23, of 18 feet lift, which connects with the pool from dam No. 13. Dayyi No. 13, 24 feet high and 600 feet long, is located 6,000 feet below lock No. 23, on firm red shale. Lock No. 24, of 10 feet lift, is on the right bank and connects with canal 550 feet long to lock No. 25, of 10 feet lift, connecting with pool from dam No. 14. Dam No. 14, 12 feet high and 6 'jO feet long, is located 6,950 feet below lock No. 25, on hard red sands one. Lock No. 26, on the right bank, is a guard-lock, giving entrance to a canal 1,000 feet long, extending to lock No. 27, of 12 feet lift, which connects with pool from dam No. 15. Dam No. 15, 20 feet high and 600 feet long, is located 11,100 feet below lock No. 27, REPORT OF THE CHIEF OF ENGINEERS. 73 G just above Buffalo Shoals, on foundation of a slaty character, overlying sandstone. Lock No. 28, of 13 feet lift, on the left bank, is at the lower end of a canal 500 feet long, arranged to dispense with a guard-lock, and connect with pool from dam No. 16. Dam No. 16, 18 feet high and 640 feet long, is located 6,500 feet below lock No. 28, on foundation of sandstone. Lock No. 29, of 15 feet lift, is on the right bank, and con- nects with the pool for dam No. 17 by a cut 600 feet long. Dam No. 17, 31 feet high and 600 feet long, is located 10,200 feet below lock No. 29; a foundation of sandstone, alternating with hard slate. Lock No. 30, of 20 feet lift, on the left bank, connects with pool from dam No. 18 by a cut 1,500 feet long. Dam No. 18, 30 feet high and 620 feet long, is located 13,500 feet below lock No. 30, near- Arbuckle Creek; foundation, sandstone. Lock No. 31, of 22 feet lift, is on the right bank, and connects with pool from dam No. 19. Dam No. 19, 21 feet high and 770 feet long, is located 16,700 feet below lock No. 31, on sandstone foundation. Lock No. 32, of 14 feet lift, is on the left bank, and connects with a canal 2,500 feet long, extending to lock No. 33, of 14 feet lift, which connects with the pool for dam No. 20. Dam No. 20, 27 feet high and 690 feet long, is located 9,000 feet below lock No. 23, about one-half mile above Sewell, on sandstone. Lock No. 34, of 22 Let lift, is on the right bank, and connects wffth pool from dam No. 21 by a cut 1,000 feet long. Datn No. 21, 31 feet high and 672 feet long, is located 12,800 feet below lock No. 34, on sandstone foundation. Lock No. 35, on the left bank, is a guard-lock, connecting with a canal 850 feet long, extending to lock No. 36, of 18 feet lift, which also connects with a canal 600 feet long, extending to lock No. 37, of 18 feet lift, which connects with pool from dam No. 22. Dam No. 22, 42 feet high and 630 feet long, is located 4,900 feet below lock No. 37, on hard slate foundation. Lock No. 38, on the left bank, is a guard-lock, having an ap- proach-cut 700 feet long above it, and connecting at lower end with a canal 500 feet long, extending to lock No. 39, of 17 feet lift, and lock No. 40, of 15 feet lift, which con- nects with pool from dam No. 23. Dam No. 23, -23 feet high and 550 feet long, is located 3,700 feet below locks Nos. 39 and 40, on sandstone foundation. Lock No. 41, of 14 feet lift, is on the right bank, and connects with pool from dam No. 24. Dam No. 24, 34 feet high and 600 feet long, is located 6,900 feet below lock No. 41, on sandstone foundation. Lock No. 42, of 19 feet lift, is on the right bank, and connects with a canal 900 feet long, extending to lock No. 43, of 19 feet lift, which connects with pool from dam No. 25. Dam No. 25, 29 feet high and 600 feet long, is located 10,300 feet below lock No. 43 ; foundation, sandstone, overlaid in places with hard slate. Lock No. 44, of 20 feet lift, is on the right bank, and connects with pool from dam No. 26. Dam No. 26, 35 feet high and 650 feet long, is located 7,800 feet below lock No. 44 ; foundation, sandstone. Lock No. 45, of 19 feet lift, is on the right bank, and connects with the pool from dam No. 27. Dam No. 27, 16 feet high and 600 feet long, is located 5,200 feet below lock No. 45, one-half mile above Hawk’s Nest ; sandstone foundation. Lock No. 46, of 19 feet lift, is on the right bank, and connects with pool from dam No. 28. Dam No. 28, 26 feet high and 600 feet long, is located 10,400 feet below lock No. 47 ; foundation on sandstone or hard conglomerate. Lock No. 48, of 13 feet lift, is on the right bank, and connects with a canal 1,000 feet long, extending to lock No. 49, of 13 feet lift, which connects with the pool from dam No. 29. Dam No. 29, 32 feet high and 600 feet long, is located 5,600 feet below lock No. 49, be- low Cotton Hill Station ; foundation, hard sandstone or conglomerate. Lock No. 50, of 20 feet lift, is on the left bank, and connects with the pool from dam No. 30. Dam No. 30, 36 feet high and 660 feet long, is located 5,000 feet below lock No. 50, on sandstone foundation. Lock No. 52, of 13 feet lift, which connects with the pool from dam No. 31. Dam No. 31, 39 feet high and 600 feet long, is located 4,800 feet below lock No. 52, just above the Blue Hole, on sandstone foundation. Locks Nos. 53 and 54, of 15 and 14 feet lifts, respectively, are on the right bank, and connect with the pool from dam No. 32. Dam No. 32, 30 feet high and 800 feet long, is located at foot of Narrow Falls, on sandstone or conglomerate foundation, 10,000 feet below lock No. 54. Lock No. 55, of 21 feet lift, is on the left bank, and connects with the pool from dam No. 33. The ap- proach to and pit for this lock are partially in solid-rock excavations. Dam No. 33, 13 feet high and 1,400 feet long, is located just above the crest of Ka- nawha E'alls, 10,200 feet below lock No 55, on foundation of sandstone conglomerate. Lock No. 56, of 15 feet lift, is on the left bank, and connects with a basin 250 feet long, extending to lock No. 57, of 15 feet lift, which connects wdth the pool below in the Kanawha River. APPENDIX V. 737 The total cost of the slack-water scheme, as shown by the appended estimates, in detail, will be — With tunnel No. 2, Stretcher’s Neck $11,427,010 With tunnel 2, 4, and 5 1 14; 102,017 And without any tunnels, about 11,000,000 The estimate No. 1, using only Stretcher’s Neck tunnel, will probably be found the better line, as combining economy of time and cost. RECONNAISSANCE OUTSIDE OF THE VALLEY OF NEW RIVER FOR TUNNEL-LINES, ETC. Examinations were made as below for determining the practicability of avoiding some portions of the valley of New River by a resort to tunnels. Instrumental sur- veys were made by Big Loup and Meadow Creek, the differences by elevation in other cases being determined by aneroid barometers, (Casella’s.) No. 1. From near Alderson’s Ferry, on the Greenbrier, by way of Griffith’s and Lick Creeks, to New River. No. 2. From the same point on the Greenbrier, by way of Little Meadow Creek, to New River. No. 8. From mouth of Piney Creek, (Stretcher’s Neck,) by way of Paint Creek, to Kanawha. No. 4. From mouth of Piney Creek to Clear Fork of Coal River, and down Coal River to the Kanawha. No. 5. From mouth of Arbuckle Creek to Big Loup Creek, and down that to the Kanawha. No. 6. From Arbuckle Creek to Paint Creek, and down that to the Kanawha. As none of these lines had any source of water-supply for a summit-level, tunnels were a necessity. Line No. 1 would require a tunnel from 6 to 8 miles long. Line No. 2, a tunnel Ilf miles long. Line No. 3, a tunnel 13f miles long. Line No. 4, a tunnel 17 miles long. Line No. 5, a tunnel 11 miles long. Line No. fi, a tunnel 12 miles long. These results were considered prohibitory, and further instrumental investigations were thought unnecessary. It is possible that lines Nos. 1 and 2 could be supplied with water by a reservoir on the site of that proposed by Mr. Ellet, on Meadow River, but at a very considerable increase of cost over the river-line, with very doubtful re- sults as to economy of time. Though the exhibit appears so unfavorable for any attempt to break away from the valley of New River with a water-line, I am inclined to the belief that the effort to accomplish the same object by a line of railway would show a more favorable result, especially for a narrow-gauge line. The whole region of country about the head of Coal River and Loup and Paint Creeks, besides being a rich agricultural country in the valleys, has in iis mountain-ranges large deposits of cannel and other coals which the configuration of the country will permit to be put cheaply and quickly in communication with the Kanawha River by means of tram-roads similar to those recently introduced into the mining regions of Spain, and advocated in Van Nostrand’s Magazine, by Mlf Herman Haupt, civil en- gineer. None of these streams, except Coal River, offer any possible facilities for water-trans- portation ; the latter stream is now “ slack-watered” to Peytona, and this work prob- ably represents the ultimate capacity of the district in that respect. In connection with the slack-water project, there is herewith submitted the report and estimates of Mr. J. M. Harris, superintendent James River and Kanawha Canal Company, to whom I am indebted for much valuable assistance during the progress of the surveys. CANAL-SURVEY. In order that the data procured in the field might enable any combination of canal and slack-water to be made hereafter, that may be considered desirable, independent of the views of those engaged in the field, two entirely distinct surveys were carried on, one all “ slack-water,” the other all “canal.” For the canal an experimental line with cross-sections of the surface at every 100 feet was traced from the last dam of the Greenbrier division above Hinton to the pool above Kanawha Falls, a distance of about 60 miles. The line was not prolonged beyond this point, for the reason that the New River for the 2 miles below to Kanawha Falls offers equal facilities for slack-water navigation 47 E 738 REPOUT OF THE CHIEF OF EEGINEERS. with the Kauawha below the falls, for which a modified form of that method of improve- ment has been definitely adopted. The location of the Chesapeake and Ohio Railroad on the ri^ht bank of the river from Hinton to Hawk’s Nest, and on the left bank below that point to the falls, renders it generally a matter of necessity that a lateral canal should occupy the opposite bank of the river, as it is in only one or two places, and for short distances, that there is sufficient space between tUe road and the river for a canal even of much smaller section than fhat considered requisite for this line. As traced and marked in the Held, the line starts from the last dam on the Green- brier, follows the right bank through the Hinton bottoms over the same route de- scribed for the “slack-water,” crosses the river at the same j)oint, and thence follows the left bank miles, where it crosses the upper end of Meadow Creek bottoms on the right bank; passing through these bottoms for about 3 miles, it recrosses the river to the left bank, which it follows 13 miles to the upper side of Stretcher’s Neck. Here the line crosses to the right bank, passes through the bend by means of a tunnel, and immediately recrosses to the left bank, which it follows 29 miles to a point just below the Hawk’s Nest, where it crosses to the right bank, and occupies that side to the ter- minus below Narrow Falls, (about 2 miles above Kanawha Falls.) From this brief description it will be seen that, as projected and traced in the field, the line crosses the river six times, viz, at Hinton Meadow Creek twice. Stretcher’s Neck twice, and below Plawk’s Nest. Of these, the crossing at Hinton is justified by the more favorable nature of the ground on the left bank, the railroad occupying the right bank ; the crossings at Meadow Creek were made with a view of utilizing the favorable ground extending along the right bank for several miles between the railroad and the river, but after-smdy would indicate that for a really independent canal this line would be inadmissible, owing to the cost of aqueducts at the upper and lower crossings. The canal-line, adhering to the left bank, between the crossing-points, would cost $768,300, and between the same points, by crossing, &c., as in the field, $800,335.25; so that without aqueducts, and with crossings in pools, nothing is gained by the crossing, and it is suggested that this crossing be abandoned. The crossings at Stretch- er’s Neck would probably have to be made, even though the line were continued around the bend, owing to the topographical features of the valley, and, as by a tunnel about 1,300 feet long a saving of distance of 3 miles is effected, the propriety of adhering to this arrangement seems evident. The crossing below Hawk’s Nest is imperative, as the location of the railroad precludes the possibility of locating a canal along the left bank below this point. There will then remain four crossings of New River that may be considered as forming necessary features in the project, which may be possibly altered as to exact location, but yet will remain to be made somewhere along the line. The method of crossing just below Hinton is in a measure fixed by the fact that the canal-bottom at that place could not readily be located high enough to enable an aque- duct to be used, which the great width of the river, moreover, would render very costly. It is therefore only proposed to cross at this place in a pool, using wire ropes with traveling-blocks for passing boats during freshets, or without steam-power. The other crossings can be made either by aqueducts or in pools, the former, of course, being the more expensive, yet it may be considered the only proper means to be used for a really independent canal. The only motive for incurring the extra cost of constructing a canal where a slack- water navigation is practicable, would seem to be to secure certain immunity from the danger of interruption of the navigation by floods, and if the continuity of this pro- tection is broken at several points, the canal, for all practical purposes, becomes nearly, if not quite, reduced to the same grade as the slack-water in this regard. However this may be, climates are submitted for making the crossings b.oth by aqueduct and pools, except at the lower end of Stretcher’s Neck tunnel, where the Avant of space for locks renders it practically essential to cross by aqueduct in order to overcome a portion of the lift on the left or opposite bank from the tunnel. The crossing below Hawk’s Nest, being in the gorge, it is suggested, also possesses claims in favor of the use of the aqueduct, arising from the declivity and contraction of the valley in that vicinity. TUNNELS. The same tunnel-lines described in the project for slack water navigation have been connected with the canal survey and estimates of cost prepared for the different sec- tions. The one similar to that proposed for the slack-water, which would permit boats to pass each other freely, is 54 feet wide at water-line. The other is similar to the sec- tion proposed for the summit tunnel, and is about 30 feet wide at water-surface. As all of these tunnels would undoubtedly (the rock being exposed along the river above and below their levels) be located in a compact sandstone, which would permit their easy excavation and maintenance with any desired width, the larger section would be well wort.h the extra cost over the smaller tunnel. Assuming that the rate of travel in the tunnels would be one-half of that in the open canal, there would be no gain of time by the adoption of any of the tunnels, except that at Stretcher’s Neck, where it might amount to forty minutes. APPENDIX V. 739 In regard to cost, tunnel No. 1, through the bend above Stretcher’s Neck, including the aqueduct, to Stretcher’s Neck, would cost about $1,000,000; and by canal and aque- duct, $1,300,000; or by canal crossing in pool, $700,000. At this point, if the plan of crossing in all cases by aqueduct were adopted, ques- tions of facility of operation, directness, &c., might justify the use of the tunnel ; oth- erwise there can be no question as to the propriety of adhering to the river at a saving of nearly 40 per cent, in cost. Tunnel No. 2, through Stretcher’s Neck, (saving 3 miles by a tunnel 1,300 feet long,) appeared to be so evidently the proper line that the surveys were not carried around the bend. Tunnel No. 3, about 4 miles below Stretcher’s Neck, would cost about $1,260,000 ; and the canal around, $656,000, or about one-half the cost of the tunnel. Tunnel No. 4, below Plawk’s Nest, including aqueducts of approach in both cases, would cost about $1,555,500 ; and by canal around, $846,000, or a little more than one- half the cost of tunnel. Tunnel No. 5, through the Blue Hole Bend, would cost about $1,330,000; and the canal around, $1,161,000 ; so that neither for economy of time nor money can any of the tunnel-lines, with the larger section of tunnel, be recommended, except No. 2 through Stretcher’s Neck. It is possible, however, that the inestimable difficulties of construction will justify the adoption of tunnel No. 5, as its cost is only slightly greater than the line around. Should the smaller section of tunnel be adopted, it would reduce the figures in the case of tunnel No. 1 by $460,000; No. 3 by $.507,000; No. 4 by $369,000; No. 5, $554,000 ; which would make tunnels Nos. 3, 4, and 5 either equal to or less than the canal-liiie around in cost, and their adoption would therefore be justifiable. In the appendix will be found a tabular statement showing distances by canal and by tunnels, and saving in distance and cost of each route. AQUEDUCTS AND CULVERTS. The aqueducts and culverts over 20 feet span are, with one or two exceptions, pro- posed to be built of wrought-iron trusses, carrying plate-iron troughs 30 feet wide in the clear. This plan was adopted on account of the difficulty of crossing either the river or its tributaries at a sufficient elevation above low-water surface to permit the discharge of flood-waters under masoniy structures. It is proposed to convey feed- water over all the aqueducts in pipes, secured to the outer ends of floor-beams, in order that the levels may be kept full when the aqueducts are occupied by boats and others are being locked through at the lower end of the levels. Owing to the very limited area of the water-shed of New River between the Green- brier and Kanawha Falls, the number of tributaries of any considerable size is unusually small, and in consequence the number of large culverts is equally limited, there being (inclusive of Meadow Creek) only 7 over 20 feet span on the whole division 60 miles in length, and only 6 if the crossing at head of Meadow Creek Bottom be abandoned. It is proposed to cross Meadow Glade and Loup Creeks by wrought-iron riveted trusses, of 66 feet, 158 feet, and 108 feet span, respectively, each carrying rectangular troughs of boiler-iron 30 feet wide in the clear. At Mill, Arbuckle, Rush, and Wolf Creek, it is proposed to use masonry arches, single spans of 40 feet, 50 feet, 40 feet, and 100 feet respectively. The aqueducts for river-crossings are three in number, all proposed to be wrought- iron Warren girders, carrying a similar plate-iron trunk to that described for the culverts. Aqueduct No. 1, upper side of Stretcher’s Neck, has 4 spans of 145 feet each. No. 2, lower side of Stretcher’s Neck, has 2 spans of 144 feet each. No. 3, below Hawk’s Nest, has, on the tunnel-line, 2 spans of 167 feet each, and on the line around the bend 3 spans of 167 feet each, and one of 100 feet. Diagrams of these trusses, with a general section of water-trough, will be found with the drawings submitted. LOCKS. Investigation of the subject since the previous reports on this work were made tending to show that a change in the proportions of the lock-chamber might be de- sirable, they have been estimated for in this project as 4 feet wider than those formerly j)roposed, making them now 120 feet by 24 feet in the chamber. Owing to tfie character of the country and the variable rates of fall in the river, it has been found difficult to establish ornidhere to any uniform lift for the locks, though the endeavor has been made to introduce as few changes as possible, and to group the yarious lifts together in sections. In j)oint of fact, the usual and ordinarily correct method has here been necessarily reversed, and the lifts of the locks increased as we descend the river; a matter of small moment in this case, however, as the river along- side will furnish always an abundant supply of water; and it is proposed to feed the levels by culverts or pipes passing around and outside of the locks. It was proposed to use 10 and 15 feet lifts as the standard, and of the 58 lift-locks, 19 are of 10 feet lift 740 REPORT OF THE CHIEF OF ENGINEERS. and 27 of 15 feet lift ; of the remaining twelve, seven have 12 fret lifts, one has 11 feet lift, three have 8 feet lift, and one has 7 feet lift. It is pro))osed to lill and empty the locks by means of culverts in the side- walls passing around the hollow quoins, as well as by slide-valves in the gates. The gates are proposed to be of wood, braced with iron ; the masonry to be similar to that proposed for the Kanawha locks. The division comprises 57 levels, their average length being a little over 1 mile ; the longest is 20,000 feet and the shortest 400 feet. CANAL-PRISM. The section adopted for canal-prism is generally 74 feet at top, 60 feet at bottom, and 7 feet deep. River-side (or tow-path) bank, 10 feet on top ; berme-bank, 8 feet on toi5 ; interior slopes, 1 to 1. Through the Hinton bottoms the prism was estimated to be 102 feet wide, in order to render this section to some extent available as a transfer-basin for the trade that will at this point undoubtedly pass up and down New River, above the mouth of the Greenbrier. An estimate is also submitted for reducing the width through the gorgre (from Sewell down) to 60 feet ; but it would hardly appear that the economy (i$500, 000 in 17 miles) would balance the loss of capacity for transportation. The surface of water in canal as located is generally kept well above the highest flood-marks, and where necessity compels any deviation from this rule, the river bank and wall is proposed to be raised to a sufficient height to prevent injury by floods. The horizontal allignment has been made with a view of using no less radius of curv- ature than 1,000 feet. In a few places this has been reduced to 800 feet where the total angular change was slight, and in others the canal has been wideyed to form a basin in which the change of direction could readily be made. The imperative necessity for preserving unobstructed the natural water-way of the river during floods, more especially in its lower y)ortions, has ]»recluded the extension of embankment-slopes into the river, and rendered necessary the frequent and contin- uous resort to retaining-walls, which will be found to form the main item of expense throughout the section. The general character of about two-thirds of the excavations on this division would be classed as loose rock, the hill-sides being unive’^sally composed of broken fragments of all sizes thrown from the adjacent cliffs, and but slightly intermixed with soil or earth of any kind. This material may naturally be expected to slide upon very slight provocation, and consequently, where deep excavations are required, the hill-slope must alw^ays be pro- tected by a retainiug-wall, which forms another la^ge item in the estimates. This same material will form the canal-prism throughout a large proportion of the whole line, (exactly how great a portion cannot be stated now,) and even with the liberal allowance of water to he had it will require some stanching to render it suffi- ciently water-tight for use. Whether any material wffiich will prove effective for this purpose can be flmnd on the highlands bounding the river valley, so as to be economi- cally applied, cannot now be determined ; the valley itself certainly offers no material in any considerable quantities at all suitable. The proportion of tbe line requiring it, and the degree or amount of stanching being necessarily indeterminate, the estimates (to be on the safe side) contemplate for the whole line a continuous lining of the wetted perimeter of the canal-prism, with 6 inches thickness of concrete, costing $26,400 per mile. It is altogether probable that one-third of this amount could be stricken off, aud_pos- sihle that it might be reduced to one-half of the whole length of the division. In preparing estimates for walls, excavations, &c., the position of the rock-line was assumed from its position, w'here visible, its general trend, &c., and in all cases of doubt the assumptions were made certainly against the canal-line. WATER SUPPLY AND FEEDERS. The low-water discharge of New River is not accurately known ; the lowest observed quantity that I am aware of was about 2,000 cubic feet per second. But the low-water discharge of the Kanawha River, which must be greater than that of New River alone, has been stated at 1,300 cubic feet per second, so that it would not be safe to estimate New River at over 800 cubic feet per second. An amount, however, that it requires no lengthy calculation to show is an abundant supply for any trade that could be passed through the projected canal-locks. It is proposed to feed the canal from the riv%r at three points in the 60 miles, about equidistant from each other, making the length of canal to be supplied by each one about 20 miles. Feeder No. 1 would be the pool in which the crossing is made from the Hinton bot- toms, and is formed by slack-water dam No. 1. Feeder No. 2 would be near Glade Creek, the pool being formed by a dam situated about half-way between dams Nos. 9 and 10 of the slack- water scheme. The length of feeder-drain is about 1,500 feet. Feeder No. 3 would be below Arbuckle Creek. The reservoir is here formed by a dam APPENDIX V. 741 located I he same as dam No. 18, (slack- water,) with some changes of dimensions, &c. The length of feeder-drain here required is 1,725 feet. Eacli leeder, supposing tbe loss from all causes to he at the rate of 100 cubic feet per mile per minute, would he required to supply about .34 cubic feet per second. It is pro])osed to form the feeder-channel by first clearing away the loosest lUhris from the hill-sides, and then covering the area required with a coarse concrete, (afterward plastered,) upon which the sidewalks would be built, to form a rectangular water-way of the requisite dimensions. It is believed that the masses of loose rock forming these si >pes will afford a sufidcently- firm foundation for such a structure, and the very great cost of excavating to solid rock will be avoided. ESTIMATES. The estimates show that for the whole line — 1. Using all the tunnels, 54 feet wide, and making all crossings of river by aqueduct, except at Hintou, (Meadow Creek crossing being aban- doned,) the total cost would be $21,255,590 2. Using 54-feet tunnels, 1, 2, 4, and 5, and aqueduct-crossings 20, 650, 895 3. Using 54-feet tunnels, 1 and 2, and aqueduct-crossings 19,762,498 4. Using 30-feet tunnels, I, 2, 3, 4, and 5, and aqueduct-crossings 19, 107,518 5. Using 54-feet tunnels, 2, 4, and 5, and crossing in pools 18, 504, 440 6. Using 54-feet tunnel, 2, and crossings in pools, (about) 17, 600, 000 Estimate No. 2, amounting to $20,650,895, is the arrangement I would recommend for an independent lateral canal, which could be reduced to $20,122,641 by making canal only 60 feet y^ide at water-surface from Sewell down. Accompanying this will be found tabulated the following detailed estimates : 1. Fifty-seven sheets of estimates of 57 levels of canal-line. 2. Eight sheets of estimates on main and alternate canal-line. 3. Two sheets of estimates for feeders. 4. One. sheet comparison of cost of tunnels of large and small section, (canal.) 5. Two sheets estimate of excavation below Sewell, if canal is made 60 feet wide. 6. Three sheets tabulated statement of cost of each level, accumulated totals, and other information relating thereto. 7. Thirty-four sheets of estimates of slack-water line. 8. Two sheets of estimates on main and alternate lines, (slack- water.) 9. One sheet table of aqueducts, giving location, dimensions, &c., (canal-line.) 10. One sheet comparative table of cost and distances by main and alternate lines. Also, the following drawings : 1. Fourteen sheets general map of division, scale 200 feet to 1 inch, showing topog- raphy, &c., and location of canal and slack-water projects ; canal in brown, slack- water in blue. 2. One sheet showing lines of exploration, (tracing.) 3. Four sheets of profile of canal-line, (tracings.) 4. Five sheets of profile of slack-water line, (tracings.) 5. One sheet general cross-section of large and small section of tunnel, (tracing.) 6. One sheet canal-lock drawings, (tracings.) 7. Two sheets slack- water-lock drawings, (tracings.) 8. Two sheets (cross-section paper) diagrams of aqueducts over Glade Meadow, Mill, Loup, Arbuckle, Rush, and Wolf Creeks, (canal-line.) 9. Nineteen sheets (cross-section paper) details of tunnel-lines and aqueducts across New River. 10. Seven hundred and forty-five sheets (cross-section paper) showing construction of canal-line. 11. Fourteen sheets (cross-section paper) with calculations of quantities and dia- grams of .33 slack-water dams. 12. Eleven sheets (crois-secrion paper) with calculations of quantities and diagrams of slack-water canals and locks. 13. Fifteen sheets calculations and details of slack-water tunnels. Note-hooks. 1. Five transit-books, (New River survey.) 2. Three topography books. 3. Eight level-books. 4. Seven cross-section books. 5. One sounding-book. 6. Also the following books, containing notes of reconnoissance of Loup Creek : one transit-book, one level-book, one barometric level. Respectfully submitted. . N. H. Hutton, Assis t an t Engl n eer. Col. Wm. P. Craighill, Major Corps of Engineers, U. S. A. 742 REPORT OF THE CHIEF OF ENGINEERS. APPROXIMATE ESTIMATE OP EFFECT OF DAMS IN GORGE OF NEW RIVER, BETWEEN KEENEY^S AND NARROW FALLS. For discharge over dams, Francis’s formula is used, as being a mean of those of other observers : R=3.33 L (/i) f. The. following assumptions are made, all of which are justified either by ascertained facts in the case of this river, conditions proposed to be obtained by theiilans for this work, or experience on other rivers of somewhat similar characteristics: 1st. That the river falls about 1 foot in 300 feet. 2d. That the bed and valley of the river is cleared of all obstructions to a chanuel- Way 280 feet wide at low-water surface, and side slopes not less than 1 to 1 for 30 feet above low-water level. 3d. That the dams are 600 feet long on comb and 30 feet above river-bottom. 4th. That at about the length of dam, behind or above it, the river-valley, at level of crest of dam, has at least a width of 400 feet. 5th. That the water in the pools will rise about 2 feet for every increase of 1 foot in depth above comb of dam and just behind it. Dam 600 /ee/ Io7)g,30 feet high; river above least width of AQO feet; river helotv least width of 280 feet. Depth on comb. Discharge in cubic feet. Area 600 feet below dam. Area 600 feet above dam. Velocity 600 feet below dam. Velocity 600 feet above dam. Remarks. Feet. 1 1,990 Sq.ft. 2, 524 Sq.ft. 12, 060 Ft. per sec. 0. 78 Ft. per see. 0.16 2 5, 640 3, 096 12, .524 1. 82 0. 45 3 10, 390 3, 676 12, 992 2. 82 0. 80 4 13, 850 4, 264 13, 464 3. 24 1. 02 5 22, 340 4, 860 13, 940 4.60 1. 61 6 29, 370 5, 464 14, 420 5. 37 2. 03 7 37, 000 6, 076 15, 904 *6. 00 2. 33 *6 feet per second— 4 miles per hour — 8 45, 210 6, 696 15, 392 6. 75 2. 94 limit of navigation. 9 53, 946 7, 324 15, 884 7. 36 3. 39 10 63, 270 7, 960 16, 380 8. 00 3. 86 11 72, 880 8, 704 16, 880 8. 37 4. 31 12 83, 000 9, 256 17, 384 9. 00 4. 77 13 93, 640 9,916 17, 892 9. 44 5. 23 14 104, 630 1 10, 584 18, 404 10. 00 t5. 68 tlsTearly 4 miles per hour above dam. REPORT OF MR. J. M. HARRIS, ASSISTANT ENGINEER. Baltimore, Md., December 9, 1874. Sir: Herewith I submit to you the report of notes taken by me at your request^ to arrive at an approximate estimate of the cost of removing the obstructions in the canon of New River, so as to render its bed suitable for slack-water navigation. These notes were commenced at a point about three miles below Bowyer’s Ferry, where the canal terminates, after passing around Keeney’s Falls. From this point downward the river in many places becomes very much contracted by immense bowl- ders of all shapes on its banks and in its bed. At other points bars have been formed by the creeks emptying material into the stream, also consisting of bowlders, and vary- ing in size from many yards down to gravel. The object of the estimate was to ascertain w'hat it would cost 'to remove these ob- structions to the natural flow of the stream, and thus cause the channel at low-water mark to be wider, to create a wider space above the bed of the stream for the water to escape in times of freshets, and thus render the bed of the stream safe for slack- water navigation. I , deem it unnecessary to mention the method of arriving at quan- tities, as the note-books will fully explain, and about which you were consulted pre- vious to entering on this duty. I will simply add that I have exercised the best judgment of which I was capable in making this estimate, and though much of the material was not actually measured, being inaccessible in a rapid river, yet, by compar- ing quantities which were actually measured with such of a similar character that could not be measured, I have arrived at an estimate whieh may be relied on as an approximation near the truth as to quantities. These quantities have been classi fled into two parts, the first composed of solid rock and large bowlders, requiring blast APPENDIX V. 743 ing. The second consisted of small bowlders, less than a cubic yard in size, loose rock and gravel. The first quantity I have found, agreeably to my estimate, to be 357,548^ yards, and the latter 35,578 cubic yards. The former at 80 cents and the latter at 60 cents per yard, produce the sum of $307,385.80, the whole amount of the estimate. You will understand that this estimate embraces all the material to be blasted or ex- cavated from the banks and bed of New River, extending from Keeney’s Falls to the Falls of Kanawha, as if there were to be slack-water throughout. But there will be several short sections of canal in this space, viz : 1,100 feet below dam No. 23, 2,500 feet below dam No. 24, and 3,700 feet below dam No. 29. The amount estimated for improving the bed of the river corresponding to these spaces is $24,337.60, which, if deducted, would leave $283,048.20 as the estimate of clearing the river of obstructions, and if the tunnel commencing opposite Pope’s Nose (railroad) tunnel, and terminating at or below the Blue Flole, is adopted, there would be another item of $45,888 to be deducted from the amount for clearing the bed of the river, leaving $237,160.20, or rather less than $15,000 per mile. I deem it proper further to state that I estimated the width of the river at two or three of tlie narrowest points which could be found, with the view to ascertain what would be the narrowest point at low-water, when the bowlders and projecting points are cut off, agreeably to the i)lan proposed and embraced in this estimate. The nar- rowest point I found would be 265 feet at low-water when the obstructions are re- moved. The next narrowest point I found to be 277.7 feet w'hen obstructions are re- moved. These points alluded to were between Cotton Hill and the Blue Hole, one at 3007 + 58, and the other at station 3063. At the former, the railroad was 64 feet from the edge of the river, but from the river’s edge on the left bank the rock rises nearly perpendicular. There would be no difficulty in sloping the rock 1 to 1 by bla.sting off a portion of it, and the same slope or greater could be given on the opposite side, so that at a rise of 30 feet there would be at least an area of 9,000 square feet to pass the floods of New River. The railroad-banks slope about to 1 at the other point, and, though the low-water surface US narrower than at the upper point, there would be less trouble in giving enough space here to pass the floods of New River than at the upper point. I saw no other points along New River where it seemed to be so much confined for want of space as at the points just named, and it was on this account 1 thought proper to ascertain its width, or at least what would be its width at low water, when the bed of the stream is prepared for slack-wuiter navigation. Respectfully submitted, by your obedient servant, J. M. Harris, Assistant Engineer. N. H. Hutton, Assistant Engineer. REPORT AND ESTIMATES ON THE IMPROVEMENT OF THE GREAT KANAWHA RIVER, BY MR. A. M. SCOTT, ASSISTANT ENGINEER. Charleston, W. Va., January 29, 1875. Colonel: I have the honor to submit the following report and estimates on the im- provement of the Great Kanawha River : Your instructions, dated August 17, 1874, and afterward somewhat modified, directed me to make such additional surveys as were necessary to enable estimates to be made, on the following plans of improving the river, so as to afford a useful depth of not less than 6i feet, or an actual depth of 7 feet at all seasons : 1st. For a lock and dam improvement from the Great Falls to the foot of Paint Creek Shoal, and for sluice-navigation in the remainder of the river, assisted by a reservoir. This to be a revision of Mr. E. Lorraine’s estimate, submitted to you in December, 1872. 2d. For a lock and dam improvement throughout, with locks about 250 by 50 feet in the chamber. 3d. For movable dams in the lower part of the river to accommodate the coal-trade, on theydan recommended for the Ohio. I will briefly refer to previous surveys and the data at hand when your instructions were received. SURVEY OF 1838, made by Mr. Charles Ellet, jr., for the James River and Kanawha Company, and em- bracing the whole river from Great Falls to the mouth. The fall and distances, as established by it, have been proved very reliable. SURVEY OF 1856-’57-’58, made for the same company under directions of Mr. Lorraine, by Mr. John A. Byers It began at the head of Huddleston’s Island, 6.80 miles below the falls, and extended 744 REPORT OF THE CHIEF OF ENGINEERS. to the mouth of the river. This survey was without'doubt very intelligently and care- fully made, and from it we have a good hydrographic map of the river, except in some of the deep pools, where hut few soundings were taken. The profile is continuous, and agrees closely with that of 1838. GOVERNMENT SURVEYS. First, a profile showing fall from Great Falls to the foot of Lykens’ Shoal, from a survey made under your direction in 1872 ; second, special surveys made under Colonel Merrill, Corps of Engineers, in 1873. The latter embrace a little more than six miles of the river at different points where improvements have been proposed or carried on. These surveys tend to prove the reliability of the old maps and profiles. To carry out your instructions it was necessary to make additional surveys at several points, particularly of localities where locks and dams were proposed, in order to select the sites as nearly as possible and make such examinations as were necessary to form an estimate. Accordingly, a party was organized, and the survey begau at the head of Loup Creek Shoal, 3.40 miles below the falls, on September 14, 1874. About six weeks were occu- pied in the field-work, surveys being made at sixteen different points. The soundings were all instrumentally located, and at each proposed site careful cross-sections run and drillings made to determine the necessary character of foundations. During the first week of the survey the water fell to a point two-tenths of afoot below what has been considered ordinary low- water mark, and two parties were started to establish references at all desirable points along the river. This was fortunately accomplished before the v/ater rose, and enabled us to reduce all work to a uniform and satisfactory reference. These bench-marks, and others made during the survey, were well established and described. In connection with these surveys and the estimates and drawings presented, allow me to refer to the valuable assistance rendered by Mr. C. K. McDermott and Mr. John S. Hogue, civil engineers. The general features of the river and the history and description of various plans that have been recommended for its improvement have been so fully presented in re- cent reports of the Chief of Engineers that nothing need be added here. Particular reference is made to your reports for 1871 and 1873, particularly to Mr. E. Lorraine’s, accompanying the latter, (Appendix T 28,) and to Colonel Merrill’s report for 1873, (Appendix M 3.) The following table, showing the fall and distance from the Great Falls of places to be mentioned in this report, and of principal points along the river, may be u,Beful : Places. Distance from Great Falls in miles. Fall from Great Falls Basin in feet. Length and fall of principal shoals, &c., in low water. Foot of Great Falls Foot of Long Shoal 00. 00 1.38 00.00 10. 38 Shoal falls 10'.28 in 4,207 feet. Foot of Loup Creek Shoal 4. 71 22.15 Shoal falls 10M2 in 8,736 feet. Foot of Lykens’ Shoal, town of Cannelton 9. 19 32.10 Shoal falls 6'. 19 in 2,250 feet. Foot of Harvey’s Shoal 10. 61 36. 99 Shoal falls 3'. 98 in 1,400 feet. Foot of Hunter’s Shoal 11.28 38. 09 Shoal falls P.60 in 950 feet. Foot of Windsor Shoal 12. 39 39. 50 Shoal falls O'. 83 in 1,300 feet. Foot of Paint Creek Shoal 15. 12 45. 70 Shoal falls 5'. 12 in 2,300 feet. Foot of Cabin Creek Shoal 20. 83 53. 21 Shoal falls 5'. 15 in 2,376 feet. Foot of Witcher’s Creek Shoal 23. 94 57. 48 Shoal Jails 3'.70 in 2,500 feet. Foot of Cat-fish Shoal, head of Charlestoji Pool, near 26. 49 60. 29 Shoal falls 1'.59 in 1,400 feet. Brownstown. City of Charleston, near foot of Charles’ Pool 36.82 60. 76 Pool 10.65 miles long, fall 0'.47. Shoal falls 2'.71 in 1,700 feet. Foot of Elk Shoal : 37. 46 63. 47 Foot of Two-Mile Shoal 39. 33 66. 55 Shoal falls 2'. 79 in 1,900 feet. Foot of Island Shoal 40. 05 68. 85 Shoal falls 2'. 21 in 1,900 feet. Foot of Tyler Shoal 41.62 72. 93 Shoal falls 4'. 10 in 5,700 feet. Foot of New Comer Shoal 43. 52 74. 96 Shoal falls O'. 59 in 500 feet. Foot of Johnson Shoal 53. 03 83r 03 Shoal tails 4'.35 in 5,600 feet. Foot of Tacket Shoal 55. 81 86. 11 Shoal falls 2' 20 in 2,904 feet. Foot of Bed-House Shoal 62. 14 89. 86 Shoal falls 2'.84 in 1,850 feet. Foot of Gillespie’s Ripple 67. 25 93. 67 Ripple falls O'. 90 in 2,200 feet. Foot of Knob Shoal 71. 78 97. 17 Shoal falls 2'.60 in 3,200 feet. Foot of Buffalo Shoal 73. 10 98. 54 Shoal falls 0' 90 in 1,000 feet. Foot of Five Ripples— Ripple, Dehby, Intermediate, 76. 34 102. 25 Total fall 2'.80 in 12,100 feet. Eighteen-Mile, and Ripple. Foot of Arbuckle Shoal 79. 20 lO.'^. 02 Shoal falls 2' 03 in 3,300 feet. Foot of Thirteen-Mile Shoal 82. 63 106. 77 Shoal falls O'. 86 in l,2o0 feet. Foot of Three-Mile or Cantrell’s Bar Point Pleasant, mouth of Kanawha o o 107. 70 107. 92 APPENDIX V. 745 PLAN rOK SLUICE NAVIGATION BELOW PAINT CREEK. It appears to be conceded that a lock and dam improvetncnt is advisable for the first fifteen miles from the Great Falls — tliat is, to the foot of Paint Criiek Shoal — and there is but one other plan to be eomiiared with the same system in the remainder of the river. This is the combination of Fisk and Ellet’s plans for sluice navigation, as suggested by Mr. Lorraine. As explained in his report referred to, it consists in making the best possible use of all the water in low stages, by “ grading the river” with an elaborate system of slu ce-dams and supplying the deficiency from a reservoir. Ordinary brush and pile dams were proposed, filled in with stone and gravel, and the tops well secured by a frame-work of sipiare timber. They were to be built square across the river, with a waterway or sluice 120 feet wide on top and 94 feet at bottom ; the bottom of the sluice to be a strong crib filled and backed up with loose stone. The top of each crib was to be placed 6 inches below the one next above, consequently requiring a dam for every 6 inches of fall in the river. Mr. Lorraine states that this must be made a matter of experiment, and thinks it might be practicable to reduce the number of dams by increasing the fall from one to another to 9 and perhaps to 12 inches. His estimate, however, was based upon a 6-iuch fall, necessitating 120 dams below the foot of Paint Creek shoal. They were to be so located as to reduce the slope on the worst shoals to 2 feet per mile. With this arrangement, and the ordinary low-water discharge taken at 1,350 cubic feet per second, there would be a theoretical depth of about 3 feet and 10 inches in the sluices. To fill the waterways and make 7 feet depth, 1,970 feet additional, or a total of 3,320 cubic feet per second, would be re- quired. This result, which is considerably larger, proportionally, than given by Mr. Lorraine, was obtained by assuming a uniform channel of the dimensions proposed for the sluices, with a slope of 2 feet per mile, and using the formula. V=^8975.41 —.10889. If the discharge of the waterway is considered as from one reservoir to another, with a head of 6 inches, a liberal calculation gives nearly the same result as above. Besides the water required to fill the sluices, about 200 cubic feet per second would be needed to keep the dams submerged, making a total of 2,170 cubic feet per second, or 187,488,000 per day, to be furnished by the reservoir. The annual available contents of the Meadow River reservoir, according to the final estimate of Mr. Ellet, would be 10,722,032,640 cubic feet. This, divided by 187,488,000, gives over 57, the number of days the reser- voir would be able to maintain 7 feet navigation, when the discharge of the river was reduced, to 1,350 feet per second. We have but little reliable data to determine how much help the river would prob- ably need, there having been no regular gauge observations taken previous to August 1, 1872. The following table is made from a reliable record, and compared as near as possible to the references used by Mr. Ellet : W ater doAvn to 0.0, or or- dinary low mark. Below 4- 0'.5. Below -f I'.O. Below + 1'.5. Below -f- 2'.0. After August 1, 1872 Bays. Days. Days. Days. Days. 18 3fi 56 75 91 Entire season of 1873 00 6 28 44 65 Entire season of 1874 7 21 54 yo 115 Both 1872 and 1874 were considered low-water seasons. The discharge, computed by Mr. Ellet from observations taken just below the foot of Elk Shoal, was, for -(- 2'.07, 8,550 cubic feet per second. We have nothing very reliable from which to determine it between -f-2'.07 and low water, but it will be safe to consider it for -}-l'.5 to be 3,520 cubic feet, the liberal estimate of quantity required to fill the sluices and keep the dams submerged. The “ low-water season ” has been taken at sixty days, in previous reports, and the record since August 1, 1872, as well as general information to be had, goes to prove this a safe assumption, and that a reservoir able to keep up the supply during fifty-seven days, of what can safely be called the minimum discharge, would bo ample for any reasonable emergency. ESTIMATE OF COST. As stated, this plan contemplates locks and dams down to the foot of Faint Creek Shoal. The details for this part of the estimate will be given under that for slack- water throughout. The table submitted, showing dimensions, costs, &c., of each sluice- dam, is, as near as possible, a revision of Mr. Lorraine’s, increased to afford seven feet of 746 REPORT OF THE CHIEF OF ENGINEERS. water. Owing to the nature of the plan, this can he considered hut an approximation. Care has been taken, however, to make it large enough, and it is thought sufficient to cover all contingencies. The following is a summary for the complete improvement : For four stone locks and dams above Paint Creek Shoal, locks 280 by 50 feet in the chamber $918, 041 00 For excavations of channels and approaches to locks 95,000 00 Total to foot of Paint Creek Shoal 1, 013, 041 00 For 120 sluice-dams below Paint Creek Shoal 490,216 00 For protection of banks below Paint Creek Shoal 48, 000 00 For excavation of channels below Paint Creek Shoal 52, 200 00 Lorraine’s revised estimate for Meadow River reservoir 533, 200 00 2,136,657 00 Add 10 per cent 213,665 00 Total estimate 2, 350, 322 00 IMPROVEMENT BY LOCKS AND DAMS THROUGHOUT. Under this head your instructions direct two estimates, one for an ordinary slack- water improvement, the other for a modification of this well-known system in the lower part of the river by movable dams, on the plan proposed for the Ohio. That for the common improvement will be given first. The general character of the surveys made in September and October, 1874, under your direction, has been explained. As stated, they were mostly directed to making approximate locations and thorough surveys at the proposed sites for locks and dams. The relative arrangement as to location, lift of locks, &c., will be given in a table with a summary of the estimate. SIZE OF LOCKS. You gave me some latitude by directing the estimate to be made for locks about 250 by 50 feet in the chamber. The large-sized coal-barges are from 120 to 130 feet long, and generally 24 feet wide. To accommodate this important interest the locks should have at least an available length of about 250 feet. They were finally planned and estimated at 280 feet between quoins, and a clear width of 50 feet in the chamber. This gives an available length of 245 feet for the full width of the lock. I think this still too short, and that they should be built to afford a clear length of 260 feet, or nearly 300 between quoins. The additional length of chamber would add but little to the cost and materially increase the capacity of the locks, as they would nicely admit four barges of the size which experience has proved to be the most economical for the ship- ment of coal, (130 by 24 feet,) or three barges and a small tow-boat. CHARACTER AND DESCRIPTION OF MASONRY. As directed, the estimate was made for stone locks, dams, and abutments of the best kind of hydraulic ra\asoury ; the general design of the locks to be similar to the last built on the Monongahela River by the navigation company, and one now in progress by the Government, at Hoard’s Rock, under Colonel Merrill. The water is admitted through the upper platform and an arched miter-wall, and discharged through the lower gates. The masonry was estimated rather heavier than that in the Monongahela locks; the river and shore walls of the chamber respectively at 8 and 6 feet wide on top, with an outside batter of about 1 in 6. Around the gates and abreast of recesses these dimensions were about doubled. The dams were planned with a width at base of not less than 10 feet, or seven-tenths the height; 9 feet wide on top, and capped with a sloping course of timber and plank, well fastened to the masonry. Abutments to be carried up 1*5 feet above the crests of the dams, 6 feet wide on top, with a double batter of 1 in 12; total length efface and wings averaging about 120 feet. The estimate includes substantial guide-cribs and ice-breakers above, and dry retaining- walls below the locks, and a liberal allowance for paving and riprapping the banks at least 300 feet below the works. FOUNDATIONS. Considerable time and labor were spent to determine their necessary character. As shown by the following table, solid rock can be obtained at seven of the sites, and it is thought can be found at others, by more thorough examination, without mate- rially changing the locations selected. At the remaining five sites, foundations of piles and timber are proposed, substantially like those adopted for the Illinois River locks, at Henry and Copperas Creeks. The first was built a few years ago, and the latter is now in progress, under direction of Colonel Macomb, Corps of Engineers. Below each dam not on solid rock, a very strong pile and timber apron is proposed, APPENDIX V. 747 extending 20 feet below the dam. The plan of foundations and apron will he shown on the general drawings to accompany this report. The detailed estimates for this improvement were forwarded to you on the 9th instant. A summary is given in the following table, with location, lift of lock, length of dams, &c. : Number from Great Falls. Location. > Distance from Great Falls. Distance from mouth of riv- er. Lift of lock. Length of dam j 1 1 Character of 1 foundation, j ! Estimated cost. Miles. Miles. Feet. Feet. 1 Near head of Loup Creek Shoal 3. 34 90. 86 9. 6 660 Rock $174, 931 2 Near foot of T.nnp Creek Shoal 4. 67 89. 53 10. 4 785 Rock 213, 105 3 Foot of Lykens Shoals, at Cannelton 9. 19 85.01 13 676 Artificial.. 272, 459 4 Foot of Paint Creek Shoal 15. 12 79. OS 13 579 Artificial.. 257, 546 5 Brown stovpn, near head of Charleston Pool.. 26. 84 67. 36 13 548 Rock 219, 940 6 First below Charleston Pool, head of Island 39. 69 54. 51 7 544 Artificial.. 218, 573 Shoal. 7 Near head of Newcomer Shoal 43. 22 50. 98 6.5 563 Artificial.. 220,142 8 Between Seary and .Tnhnsnn Shoals.. 52. 46 41. 74 7 598 Rock 182, 644 9 Hea,d of Red House Shoal 61. 77 32. 43 6. 5 670 Rock 191, 067 10 Near foot of Gillespie’s Ripple ... 67. 36 26. 84 6 590 Rock 164, 187 11 Head of Hobby’s Ripple. 75. 02 19. 18 7 558 Rock 170, 340 12 Foot of Three-Mile or Cantrell’s Bar 92. 40 1. 80 8.7 697 Artificial.. 244, 442 Total for leeks and dams 2, 529, 376 Estimate for excavation of channels anti approaches to lochs 268, 000 2, 797,37) 10 Tier oent. for onutin oTfineips .. 279, 737 Total estimate 3, 077, 113 IMPROVEMENT BY MOVABLE DAMS, AS PROPOSED FOR THE OHIO RIVER. For a description of these dams, it is only necessary to refer to the elaborate report of the Board of Engineers on Movable Dams, &c., consisting of General Weitzel and Colo- nel Merrill, (Ex. Doc. No. 127, H. of R,, 43d Congress, 1st session,) and to Colonel Mer- rilPs annual report, printed as Appendix N of the Report of the Chief of Engineers for 1874. As shown in these reports, this modification of the common slack-water plan is designed particularly to accommodate the coal-trade, and you directed me to extend the estimate for it as high up the river as this interest seemed to require. In view of the importance of this interest and the great advantages of movable dams to suit it, it was thought advisable, in planning for the ordinary improvement, to do so with ref- erecce to their final adoption below Charleston. Accordingly the maximum lift for the locks in this part of the river was taken, as by Colonel Merril for the Ohio, at 7 feet. This would virtually extend the movable system to Brownstown, at the head of the Charleston Pool, and though the coal-field extends still higher, the arrangement it is thought would accommodate the whole interest very well, as there would be at most but two lockages to get the coal into the Charleston Poo), where there would he every facility for harboring it and for making up tows. If it should ever be found de- sirable to extend the movable dams as high up as Cannelton, at the foot of Lykens Shoal, it could be done by building two intermediate locks and dividing the lifts of Nos. 5 and 6. This arrangement makes the number of locks the same as proposed by Mr. Lorraine in his report to yon, and also as recommended by Mr. John A. Byers in 186H, when he made a general estimate for a lock and dam improvement below Loup Creek. It adds one to the number of locks below' Charleston, however, and reduces the number above to five. The advisability of the high darns proposed at Paint Creek and Brow'nstown, particularly at the latter place, is somewhat questionable, but they are no higher than some on the Monongahela River, where the height of banks and general characteris- tics fire similar, and as the arrangement is quite desirable I conclude, to lecommeud them. In the table the lift of lock No. 12 is given at 8'. 70 ; this is with reference to present low-water at the mouth of the river. It will finally depend on the connection made with the lock and dam system projected on the Ohio. Colonel Merrill informed me that he could/give no definite information about it, but thought there would be no difficulty in adopting their plan to suit any arrangement on the Kanawha. The low> water surface at the mouth can easily be raised enough to reduce the lift of No. 12 to 7 feet ; and if it should j)rove expedient to raise it about 3^ feet, the first lock on Ka- 748 REPORT OF THE CHIEF OF ENGINEERS. nawha can be located at tbe foot of Thirteen-Mile Shoal, where rock-foundation can be had. A summary of the estimate for the complete improvement by movable dams below and permanent dams above Charleston, with locks 280 by 50 feet, is submitted. That for the movable dams is based on the price per linear foot taken by Colonel Merrill for the Ohio, (Appendix N, Report of the Chief of Engineers for 1874.) The details are given in the same (N 9) in Lieut. F. A. Mahan’s report on the Youghiogheny River. The width of “pass” proposed for the Ohio has also been adopted, as it has been found necessary to make the present towing channels on Kanawha about 250 feet wide, to answer the requirements of the coal-trade. Number of dam from Great Falls. Number of dam from Charleston. Location of dams. Distance from Great Falls in miles. Length of pass in feet. Cost per linear foot. Estimate for pass. Length of weir in feet. Cost per linear foot. Estimate fur weir. Total length of dams. Total coat. 6 1 Head of Island Shoal 39. 69 250 $344 $86, 000 294 $227 $66, 738 544 $152, 738 7 2 Near head of New Comer Shoal. . 43. 22 ' 2.50 341 86, 000 313 227 71, 0.51 563 157, 051 8 3 Between Scary and Johnson’s 52. 46 250 344 86, 000 348 227 78, 996 598 164, 996 Shoal. 9 4 Head of Red-House Shoal 61.77 250 344 86, 000 420 227 95, 340 670 181, 340 10 5 Near foot of Gillespie’s Ripple. . . 67. 36 250 344 86, 000 340 227 77, 180 590 163, 180 11 6 ! Head of Debby’s Ripple 75. 02 250 344 86. 000 308 227 69,916 558 155, 916 1-2 1’, Foot of Three-Mile Bar 92. 40 25( 344 86, 000 447 227 101, 469 697 187, 469 602, 000 560, 690 n’ntn.l for movalde dams 1, 162, 690 Locks and abutments, as shown in estimate for slack-water throughout . . 971, 496 Total cost of movable-dam improvement below Charleston 2, 134, 186 Estimate for five locks and permanent dams above Charleston Pool . 1, 137, 981 Estimate for excavation of channels and approaches to locks 268, 000 3, 540, 167 Add 10 ner cent, for contingencies . 354, 016 Total estimate 3, 894, 183 In regard to the relative merits of the two plans of improvement, (not to consider the movable dams,) I believe the slack- water can be much more confidently recom- mended. There are elements of uncertainty connected with the plan for open navi- gation ; and it is thought, with every condition realized, it would have no advantages over locks and dams. The sluices are not more than half wide enough to answer the requirements of open navigation, and it appears they cannot be made much wider than planned, (120 and 94 feet,) if dependence is based on the one reservoir. This is about the width of the present “dug chutes” on the river, and, in stages w'hen the navigation is confined to them, experience has limited descending tows to two and never more than three loaded barges. The great difficulty anticipated, however, is to ascending craft. No form of sluice can obviate the unpleasant currents to be encoun- tered in entering and passing them, and in the night or windy weather, particularly in certain stages of water, this would be rendered more or less dangerous. Mr. W. R. Hutton, in his report to you dated January, lii71, alludes to the resistance to a tow of loaded boats passing the sluices up-stream, and says they would necessitate some modification of the present system of towing. If the tows should be limited to the size that could enter the locks proposed, (three barges and a tug) it is thought 120 sluices would cause much more trouble and delay than the eight locks proposed in the same distance. In addition to these objections, and the uncertainties of realizing every condition of a theoretical plan, there is an element of danger connected with a reservoir of the dimensions jjroposed, formed by a dam nearly 70 feet high, which should not be ignored. From past reports it appears that a jdan of open improvement has been sought for two principal reasons, one being the local objection to slack-water improvement, the other the incidental assistance which the reservoir would render to the Ohio. I think it may be said that neither of these reasons now demand consideration, for the people of Kanawha are almost without exception in favor of locks and dams, and it is gen- erally conceded that the same system, or some modification of it, must be resorted to on the Ohio. Pains have been taken to learn the sentiment of river-men, who are APPENDIX V. 74D generally familiar with this plan from observation of the Monongahel a slack-water^ and they are found universally anxious to have it adopted on the Kanawha. If the movable dams, so confidently and ably recommended for the Ohio, should prove successful, they would remove almost every possible objection to the slack-water improvement. Very respectfully, ycnr obedient servant, A. M. Scott, Assistant Engineer. Col. W. P. Cl{AIGTlILT, Major Corps of Engineers, U. S. A. REPORT ON IMPROVEMENT OF GREAT KANAWHA RIVER P.Y MEANS OF LOCKS AND DAMS,. RV MR. WM. R. HUTTON. Baltimore, Md., Jane 30, 1870. Sir : The project for the permanent improvement of the Kanawha River as recom- mended by the Board of Engineers on the ‘i5th of May, lh75, (see page 94, j)art 2, Re})ort of Chief of Engineers for 1875,) has for its object to furnish a navigable depth of 7 feet at all seasons of the year, from the mouth of the river to the falls. It pro- poses to accomplish this by means of nine locks of low lift, with movable dams, fiotu the mouth to the foot of Paint Creek Shoal, which is 15 miles below the falls, these 15 miles being improved by three locks of 15 feet lift each, connected with iiermanent darns. The project is based upon the very complete surveys made in 1858 by Mr. John A. Byers, and special preliminary surveys of the different sites selected fbr permanent works made under your direction by Mr. A. M. Scott in 1873, 1874, and 1875. In 1873 the low-water surface of each jiool was observed, and referred to a proper permanent bench-mark, the surface of Charleston Pool reading 1.50 on the gauge-board at that jdace. This has been adopted as standard low water, although in 1874 the sur- face of the pool was 0.2 foot lower. In 1875, an accurate survey was made from Cabin Creek Shoal to a point 2^ miles below Brownstown, and a line of careful levels was run from the Falls to Newcomer’s Shoal, 7 miles below Charleston. The only recorded gauges of the discharge of the river made before last year were those of Mr. Ellet. During 1875, twelve gaugings were made by Mr. Scott at the site of lock No. 5, 8| miles above Charleston, at various stages, ranging from 1.55 feet to 32 feet above low-water. The river was very full during the entire season, and at no titoe did it fall to low-water mark. The river is fully described, as to its general characteristics, by Mr. Lorraine, in his report to you of December 9, 1872, (see page 836, Report of Chief of Engineers for 1873.) I recapitulate the principal features. The length from the Falls to the Ohio River is 94 miles ; total fall from the pool at the foot of the Great Falls is 108 feet, 47 feet of which occurs in the first 15 miles. The average width is 590 feet. During low water it is navigable from the upper end of Charleston Pool to the mouth by steamers draw- ing 3 feet. Above Charleston Pool, there is no navigation at low stages, although some improvements have been made in the way of sluices and training-walls, which are valuable to navigation at moderate stages. Coal is not shipped, however, until the river rises 5 or 6 feet at Charleston. Ordinary floods rise to 25 or 30 feet above low water in the upper half of the river, which here does not overflow the general level of its banks. The highest flood on record, that of 1861, rose to 47 feet. Extreme low-water discharge, according to Mr. Ellet, is 1,100 cubic feet per second; ordinary low water, about 1,300 cubic feet. Comparing the Kanawha with the Seine, which has been improved with movable dams, the low-water discharge of the Seine is 1,700 cubic feet per second, that of the Kanawha being 1,350 (rarely 1,100) cubic feet. The maximum flood of the Seine (anno 1658) rose 29.3 feet; of the Kanawha, 47 feet in September, 1861, at Charleston above (0) of gauge, which is 45^ above low water. Ordinary high floods of the Seine 20.5 feet, with a discharge of .56,000 cubic feet ; of the Kanawha, 36 feet, discharging probably 132,000 feet. The low-water slope of the Seine above Paris is 0.5 foot per mile ; of the Kanawha below Paint Creek, 0.8 foot per mile. Although the Kanawha is not navigable at low stages, there is no flood in which boats do not run. In this it differs from the Seine, where navigation ceases when the flood exceeds 10 or 12 feet. As has been already mentioned, the project for the improvement embraces the con- struction of nine locks with movable dams, and three locks w ith permanent dams. The greater slope above the foot of Paint Creek Shoal renders it inexpedient to con- tinue the improvement by movable dams above that point. The movable dam, which has led to the recent great development of the interior 750 REPORT OF THE CHIEF OF ENGINEERS. navigation of France, is fully described and figured in tbe report of Weitzel and Mer- rill, upon “ Hydraulic Gates and Dams,” (see page 638, part 1, Report of Chief of En- gineers for 1875.) Its object is to permit an. open river navigation whenever the natural depth of the water is sufficient, and to furnish a navigation by locks and dams during low stages of the river. The navigation pass or sluice, in the present project 250 feet in width, is an opening in the dam, with its sill about the level of the bottom of the river. It is furnished with wickets, which may be raised to close it to the full height of the dam, and which lie flat upon the floor when tbe pass is open. The fixed portion of the dam is usually provided with wickets of a less heigiit, to regulate the water in the pool, and to facilitate opening and closing the wickets of the pass. Their use is not contemplated in the original project, but further study indicates the substantial advantages which would follow their introduction. Good river navigation above Charleston requires at least a stage of 6 feet on Charles- ton gauge. With this height there is a rise of about 5.5 feet at lock No. 5 ; so that with the sill of the pass placed one foot above bottom, we have 8.5 feet of water in the pass. With 6 feet on Charleston gauge, the discharge of the river is about 12,000 cubic feet, giving a velocity in the pass of about 4. .5, or a height on the dam, if closed, of 3.06 feet. As the water rises, we have, with a depth of 10 feet in the pass, a velocity of 4.8 feet ; with 12 feet, 5.8 feet per second ; and when the pass is full, but none discharging over the weir, (supposed without wickets,) 6.4 feet, or about 4 ^ miles per hour, the lock-gates also being open for the passage of the stream. To diminish these velocities, as well as to reduce the maximum height of the sheet flowing over the wickets of the pass, which will permit a reduction of their strength and weight, the dam might be finished at 4 feet below the level of the pool, and furnished with wickets of that height. These will be comparatively inexpensive; being low, they may be wide, say 6 feet at least, and they will need no slide or tripping-bar. Reference is intended to the wick- ets of the Chanoine system, which should be preferred on account of their cheapness. The Desfontaines system is more convenient, but the first cost is much greater. The surveys of the past season have enabled definite locations to be made for locks and dams Nos. 4 and 5. the former at the foot of Cabin Creek Shoal, and the latter about a mile below Brownstowu, near the head of Charleston Pool. Lands have been purchased at both sites, and both locks are under contract, as well as the dam at No. 5. The surveys have also developed the fact that the river-bed is underlaid at no great depth by a ledge of rocks lying nearly parallel to the general slope of the river. Its continuity is not perfect; but it is so general that we may reasonably expect to find suitable locations for all the works, where they may be founded upon the rock. The levels that have been taken show the changes that have occurred in the height of surface of the pools in the past twenty years. The improvement of the bars by in- creasing the water-way through them, increasing their depth, at the same time lowers the surface of the pool above it. Thus, Charleston Pool has been lowered nearly 2 feet bv th« great improvements to the channel at Elk Shoal; and the same effect is observed at other points. tr. As a type of the low lock and movable dam, a description is introduced of No. 5, at Brownstown, some eight miles above Charleston, now under construction. The lock, Avhich is 364 feet in toral length, 300 feet between hollow quoins, and 50 feet in clear width, is designed to pass at one lockage four coal-barges of the dimensions usual on the Ohio and Kanawha Rivers ; that is, 130 feet long, 24 feet wide, and drawing 6 feet of water. It is placed in the river as near to its left bank as practicable, while giving a good entrance and exit for boats. The dam, about 560 feet long, crosses the river at right angles to its direction, opposite the lower abutment, with a height from the rock of about 17 feet, and will raise the water nearly 11 feet above low water. The pass adjoins the lock, is 250 feet wide, and its floor is 50 feet long in the direction of the stream. The pier which separates the pass from the river is 13.5 feet wide, 48 feet long, and 4 feet higher than the dacn. The right bank of the river is protected by a masonry abutment rising 10 feet above the dam, with wings extending to the top of the bank, above the reach of overflow. The lock- walls will rest upon the rock, which at this point is between 3 and 4 feet lower than the miter-sills. At the ends (the abutments for the gates) they will have a thick- ness of 16 feet ; the interior faces will batter one-fourth of an inch to the foot ; the ex- terior faces will be vertical on the river side, while, the backs of the shore-wall will batter 2 inches to the foot. Between the ends or abutments, the chamber- walls are 12 feet thick at bottom of lock and 5 feet at top, the whole height being 20 feet, depth of water 7 feet, and lift of the lock 7 feet. The top of the look is therefore 6 feet above the top of the dam. The upper and lower miter-sills are placed at the same level, so that, by opening all the gates, the lock may serve as an extension of the pass. The miter-sills are of stone faced with timber ; the floor of the lock is of concrete covered with plank, «s:cept a space below each gate, which is paved with dressed stone. The angles, hollow-quoins, miter-sills, coping, &c., are of cut stone ; the rest of the masonry will be of a much cheaper class, although not inferior in fitness, strength, and durability ; all to be laid APPENDIX V. 751 in the best manner in hydraulic cement. The chamber-walls will be furnished with rings for the purpose of securing boats. The locks will be hlled and emptied through valves in the gates, in addition to iron culvert-pipes passing around the hollow quoins. It is proposed to build the gates of iron frames, covered with a sheathing of plank. In the lower abutment on the river-side is a well containing tlie gearing for tripping or throwing down the wickets of the navigation pass. This, as well as the chain wells and gearing for operating the gates, is contained below the surface of the coi)iug and covered by iron plates ; for, as the lock will be submerged in tloods, it is necessary that no machinery or framing should be exposed on top of the walls. The floor of the pass is about 3 feet above the rock foundation. It is formed of con- crete, supported front and rear by timbers framed into crib-work ; the top is furnished with large timbers arranged to hold the journals, slides, &.C., which are attached to them, and paved between the timbers with stone. The wicket is a wooden frame 13 feet 6 inches high and 3 feet 8 inches wide, covered with planks, and capable of revolving about an axis placed at the middle of its height. This axis is formed by the cross-head of an iron frame or horse, which is itself movable about a horizontal axis fixed upon the floor. When a wicket is raised, its foot rests against the sill of the floor; its horse is main- tained vertical by an iron prop, the foot of which abuts against a heurter, an iron abutting piece fastened to the floor. The wickets, being placed side by side across the current, form when raised a dam to close the pass. A movable foot-bridge is constructed up-stream, across tbe pass, to facilitate the operations of opening. To throw down or open a wicket, it is necessary to pull sidewise the foot of the prop, so that it shall slide across and clear of the foot of the heurter. The prop, having lost it support, slides upon the floor down-stream; the horse at the same time turns about its axis and falls upon the floor ; the wicket follows them, and covers and protects the other pieces. If there were no water upon the floor, the wicket would be broken by the fall, but a very shallow cushion of water is sufficient, in a great measure, to destroy the shock. The foot of the prop is pulled away from the heurter by means of the tripping-bar, one end of which carries a rack, gearing into a pinion in the well in the lock- wall, worked by a crank on top. To raise the wicket, the lock-keeper stands upon the foot-bridge, and by means of a portable winch pulls up a chain attached to the foot of the wicket. The wicket rises, maintaining, however, a position nearly horizontal. It offers, therefore, but little resistance to the current. The horse and the prop follow its ascent until the foot of the latter, passing over the inclined plane, which forms the top of the heurter, falls in front of it, at the same time that the horse attains a vertical position ; the axis of the wicket is now in its final position, and a slight push on the foot of it, or even the slack- ing of the chain, is enough to cause it to right itself and bear its foot against the sill. The three upper locks will be connected with fixed or permanent darns, and will be founded upon the rock. The upper miter-sills will rest upon breast-walls, through Avhich the filling-culverts will discharge. The dams will be of masonry, with aprons of crib-work filled with concrete to protect from scour the soft rock of the foundation. The head-walls of the lock will be carried up to a sufficient height to permit their being used in moderate floods. The locks, with movable dams, will generally be founded upon the rock. They have no breast-walls, but both ends of the lock are upon the same level, so that it may be used for purposes of navigation when the pass is open. The walls extend but 6 feet above the dam, and will be submerged in floods. When rock foundations cannot be obtained at moderate depths, both lock and dam will be built upon piles. The floor of the lock will be an inverted arch to withstand the upward pressure of the water; the floor of the pass will be a heavy bed of concrete on piling, and filtration will be prevented by rows of sheet-piling and cross-walls of concrete. The cost of completing the improvement on the present plan is estimated, after careful revision, at $4,132,500. But it is proper to add that the work now under con- tract has been let at prices very much below those used in the estimate. Kespectfully^ W.Ai. K. Hutton. Col. Wm. P. Craighill, Major of Engineers, U. S. A. COMMUNICATION TO THE riTTSlJURGII COM^MERCIAL RELATIVE TO THE OHIO RIVER IMPROVEMENT. To the Editor of the Pittsburgh Commercial : Great public improvements are rarely accomplished without violent opposition, and strange as it may seem, the most strenuous opponents are usually those who in the end 752 REPORT OF THE CHIEF OP ENGINEERS. derive the grf at^’Ht heuefit from the measures they oppose. When agricultural imple- ments began to be introduced, farm laborers, led by farmers, destroyed the machines and held public meetings to denounce whoever favored their introduction. Planing- mills were opposed by the torch of the incendiary. Canals, when projected, have been denounced by wagoners, and railroads, in their turn, by the canal-men and their associated interests. Railroads have been threatened with annihilation lest they might destroy the market for horse-feed, or carry fright and death among the farmers’ herds. It is not to be supposed that the obstructives and croakers in such cases are con- sciously instigated by bad motives; on the contrary, a majority of them sincerely be- lieve their rights and living endangered. In most cases they are blinded by the ex- travagant dtclamations of a few over-confident and mistaken leaders, whose wild assertions and groundless predictions they mistake for facts and arguments. THE MONONGAHELA NAVIGATION was built in the face of bitter opposition from the flatboat-men and farmers along shore. “It is a rt markable tact,” says one of the early reports, “that with so many unanswerable arguments to recommend it to and enforce it upon the public attention, no work in the country has ever encountered greater obstacles than this. Instead of being, as it ought to have been, fostered by our citizens, and hailed by the Mononga- hela Valley as a blessing to themselves, it met with nothing but the most chilling re- gards from the one or the most determined hostility from the other.” It was declared that the obstructions and tolls would extinguish THE COAL-TRADE and destroy the value of coal-lands, and insutferable delays and expenses caused by the locks would ruin the river tor steamers. As the result of the improvement the coal- trade has increased since its construction from 400,000 bushels to 03,707,500 bushels, and the increased value of the lands has been many times more than the entire cost of improvement. As to delays to steamers on account of the locks, Capt. E, Bennett, the most experienced boatman on the river said, “ The time of running the fifty-five miles, including the passing of the four locks, varies from five and one-quarter to six hours (9 to 10.^ miles an hour, including the delays,) by different boats,” not including stop- i)ages for freight and passengers. * ^ * -Jf # “I do not recollect coming up from Pittsburgh to Brownsville before the completion of the slack-water in le^s than twelve hours, and frequently from twenty to twenty- four.” The uniform depth and the absence of current far more than compensate for the delays of the locks. COLONEL Merrill’s plan. The plan for the improvement of the Ohio is the result of 30 years’ additional study and experience of the best engineers of the world, and is infinitely superior to the Monongahela navigation ; nevertheless, opposition is the fate of every improvement and must be encountered and overcome by this one also. We could hardly expect the owners of an $80,000 tow-boat to be pleased at first blush with an improvement which will enable a $4,000 or $5,000 tug with its barges by running safely all the year round, carrying coal down and freight up, to be a suc- cessful competitor, and this is one of the results anticipated from the improvement. A memorial to Congress, which I have seen to-day, states that only one interest of the vast commerce of the Ohio is opposed to the improvement, and that only in x)art. Having knowledge that some of the parties are in opposition only from a misunder- standing of the subject, let mo endeavor to describe the plans and manner of working as they present theniselves to my mind, and their effect upon navigation. 1. A lock is to be erected on one side of the river, the outer wall of which will be 770 feet long, parallel with the current, and forming one side of the navigation-pass. 2 . Opposite to the middle of the lock, extending out from the other side of the river, is a permanent dam or weir of a proper' height, which reaches a point 400 fe-t from the wall, leaving a channel or navigation-pass between the dam and the wall 400 feet wide and as.deep as the present river-bed. 3. The navigation-pass has a system of gates or wickets capable of being raised so as to close it when it becomes necessary to convert the navigation into slack-water. 4. In order that the permanent part of the dam may make the least possible ob- struction to the water in times of unusual floods, and that it may be used in regulating the depth in the navigable pass, it is made very low and capable of being increased in height when needed by a similar arrangement of wickets. Thus in all high and moderate stages of water there will be a clear unobstructed channel 400 feet wide. When the water has fallen below the minimum depth fixed upon for the navigation — say C feet — the wickets in the pass will be raised, and there will be until the next rise a slack-water navigation in the river. APPENDIX V. 753 The locks will be 639 by 78 feet iaside, and will pass tea coal-barges, with large tow- boat and fuel flat, at one lockage. This method has been in successfal operation ON THE SEINE and other French rivers for years without any serious accident or failure. There are navigation-passes on the Seine 230 feet wide and looks 615 by 40 fliet. The barges on the Seine are about the size of those on the Ohio, though the fleets are somewhat smaller. The Ohio being a larger stream, bearing larger fleets, the navigable passes and locks are correspondingly enlarged ; and this being the only difference from the plan in successful operation, to assert that ‘‘it will not work” is simply a waste of words. The width of the open navigable pass is only 120 feet less than that of the pass now being constructed at the mouth of the Mississippi River, and is 100 feet wider than the ch-innel between the piers of the Steubenville bridge. It is 320 feet wider than an average coal-fleet, nt arly 300 feet wider than the widest and nearly double the width of the present wing-dam channels which they will supersede. The operation of this SYSTEM UPON THE OHIO will not only leave the river free for the present method of nayigation upon floods, but it will increase the length of time during which the flood will be available. This I will explain as well as I can in the absence of drawings. On the French rivers when the water is high THE NAVIGABLE PASS is open and all the wickets on the low part of the permanent dam are down. When the water falls to the minimum depth required for the navigation there is still several feet running over the low permanent dam or weir, and the lockmen then begin to raise the wickets on the weir at the shore end. Each day they raise whatever number may be necessary to turn the additional amount of water into the navigable pass which may be needed to keep up the required depth. When all the water of the river is thus made to go through the pass, and it still continues to fall, they begin to close the main wickets day by day as fast as is necessary to keep up the depth in the navigable pass by contracting the iviclth. Thus they keep open the gap, gradually narrowing its limits, as long as it can be run with safety to the works; then close it, and use the locks dur- ing the continuance of low water. Let us see how this would affect the Ohio. I have before me, in the engineer’s report for 1871, a table showing the DAILY STAGE OF THE WATER by the Pittsburgh gauge for the year 1868, about an average year. From this it ap- pears that there was — 8 feet and over, 65 days in the year. 6 feet and over, 153 days in the year. Between 5 and 6 feet, 44 days in the year. Between 4 and 5 feet, 31 days in the year. Under 8 feet, 300 days. Under 6 feet, 212 days. Under 5 feet, 168 days. It must be evident to even a casual observer of the Ohio that, if the quantity of water which flows in the open river to make a 6-ioot stage could be concentrated into a pass only 400 feet wide, it would increase the depth in the pass to an amount pro- portionate to its volume and velocity. So also the water which makes a 5-foot stage in the open river. Applying this plain fact, we find that if by raising the wickets on the permanent part of the dam during a 6-foot stage and turning all the water into the navigable pass, and the depth of the pass would be increased only 2 feet, there would be 8 feet in the channel for 153 days instead of 6b days, as non\ If concentrating the water when there is 5 to 6 feet would raise it in the pass from 1 inch to 1 foot, we would have an open 6-foot navigation 197 days instead of 153 as now. By raising half the wickets in the navigable pass and reducing the width 200 feet asthe water begins to sink to the 4-foot stage, the 6-foot open navigation could probably be continued 31 days longer, increasing the 197 days to 228 days, and leaving but 137 days during which the locks would be necessary for navigation. Thus we would get an 48 E 754 REPORT OF THE CHIEF OF ENGINEERS. improvement from the navigable passes and dams'alone, even without the locks, rep- resented by the following figures : Tlnimproved river. Xavigrable pass. Days. Days. Depth, 8 feet and over (35 153 Depth, 6 feet and over 153 197 By narrowing the pass 228 To make these figures mathematically correct would require intricate calculations and measurements involving velocity, currents, &c., but they are accurate enough for illustration, and approach very near to what will be realized from the proposed method of improvement. They show that the Ohio can have a practically unobstructed channel for navigation 6 feet deep and over for 228 days in the year, and slack-water with convenient and capacious locks the remainder of the time, during the greater part of which it is now uunavigable. F. R. B. Pittsburgh, February 19, 1876. REPORT ON THE SURVEY OP THE SUMMIT-LEVEL, BY MR. E. LORRAINE, PRINCIPAL ASSISTANT ENGINEER OF THE JAMES RIVER AND KANAWHA CANAL. Richmond, January 20, 1852. Sir : I have the honor to submit to you the following report of the survey of the summit-level of the James River and Kanawha Canal, made under your instructions dated April 7, 1851, and in accordance with subsequent instructions received from time to time in the field and by correspondence. The object of the survey was to ascertain the practicability and cost of supplying with water the summit level of the canal. A survey for this purpose was made in the year 1827 by Capt. William G. McNeill, of the United States Topographical Engineers, by whom it was pronounced practicable to feed from the Greenbrier. As his plan, however, involved the adoption of a tunnel five miles long through the Greenbrier Mountain, it was deemed expedient to make a further examination of the Greenbrier and its tributaries to ascertain whether it was practicable to feed from the Greenbrier and at the same time dispense with the long tunnel. The survey was commenced at the eastern base of the Alleghany Mountain, where it was connected with station 1093, the terminating point of a survey of the line of canal from Covington westward, made under your direction by John E. Mills, esq., in the fall of 1850. Having established the summit-level, as directed by you, at an eievation of (396 feet above the level of Jackson’s River at Covington, and 1,916 feet above tide, a location was made for the tunnel through the Alleghany to the point wUere the plane of the summit-level intersects the western base of the mountain, in the valley of Tuck- ahoe Creek. From this point I traced a line for a feeder along the eastern side of the South and Middle Forks of Howard’s Creek, crossing over the Middle Fork to the North Fork; then crossing the North Fork and running down its western margin along the base of Greenbrier Mountain, and around the south point of that mountain to the eastern side of the Greenbrier River, and thence up the Greenbrier. This line was traced with great care for about nineteen miles up Greenbrier River, when the impracticability of feeding from Greenbrier by drawing the water from any pond that might be formed upon this level became so evident, for reasons that will be hereafter stated, that, after advisement with you, the location of the feeder-line was abandoned. The levels and compass-line, however, were continued up that river to the Droop Mountain or the mouth of Spice Run, a distance of fifty-six miles from the summit-level, and about forty miles up the Greenbrier froni the mouth of Howard’s Creek. The Greenbrier River is bordereift by mountains which generally slope down to the water’s edge, with occasionally a narrow strip of low grounds intervening between the base of the mountain and the river. The sides of the mountain have a general inclination of from 25° to 35°, and not unfrequently of 45°. From Greenbrier Bridge for 19 miles up the river they are covered with loose rock, in many places perfectly bare of soil, and from 6 to 10 feet deep, measuring vertically. A feeder-canal con- structed in such a locality would require very high walls on one side, for which it would be difficult to procure a substantial foundation, and on the other side, in order to obtain a sufficient slope, a great extent of the surface of the hill would have to be removed. The whole of the bottom and the sides up to the water-line would have to be lined with puddle, which could only be procured from a great distance, and gen- erally from the opposite side of the river, and hauled up a steep hill, upon an average of 150 feet above the river: and then after the feeder was constructed, in the most substantial manner, it would be liable to be swept away by slides of eaith, stones, and APPENDIX V. 755 trees, brought down from above by every heavy rain that occurred. The entire length of the feeder would be about 53 miles, which might be shortened to 31 miles by the adoption of Captain McNeill’s 5-mile tunnel, or to 43 miles by the adoption of a tunnel miles long through the Greenbrier Mountain, opposite the White Sulphur Springs. With either of these tuuuels the general character of the feeder would be, as above de- scribed, difficult to be made substantial, and extremely costly. Those are the consid- erations which Jed to au abandonment of the plan of supplying the summit-level by means of a feeder from the Greenbrier. Having concluded the survey of the Greenbrier, my attention was directed, by par- ticular instructions from you, to the other streams which are so abundant in that mountain region, with a view of testing their availableuess as feeders. The mosc prom- inent of these are Anthony’s Creek, Little Creek, the Middle and N<'rth Forks of How- ard’s Creek, and the South Fork of Howard’s Creek, or Tuckahoe Creek. The first in rank of these streams is Anthony’s Creek, which is a tributary of the Greenbrier. It is 25 miles long from its mouth to its source, and is fed by Little Creek, and three prin- cipal branches which unite about 15 miles from its mouth, and also innumerable smaller streams and springs which issue from the sides of the mountains by which it is bound. About 8 miles above its mouth there is a place called the “ Narrows,” where the creek has forced its way through a steep and narrow gorge of the mountain, and above which the mountains diverge, and the sr.ream runs through a beautiful valley about a half-mile wide. This place was selected as a site for a mound, which, when thrown across this narrow point, will effectually arrest the water that flows down the creek, and convert the valley above into a magnificent reservoir 9 miles long and 40 miles around, with an average width of one-half of a mile, a superficial area of 2,753 acres, and a mean depth of 60 feet. The mound for this reservoir will be 126 feet high and 395 feet long. In order to avail ourselves of this immense body of water, it will be necessary that the mountain-ridge which separates the southern border of the reservoir from the valley of Howard’s Creek shall be pierced by a tunnel 2 } miles long. The level of the bottom of this tunnel wdl be 30 feet below the surface of the water in the reservoir. It passes for its entire length through a black-slate rock of easy excavation, and as it will only be necessary to be made just large enough to be advantageously worked, it cannot ^e considered an obstacle of any serious importance. After passing through this tunnel, the water from the reservoir will flow down the bed of Dry Creek, and at the narrow gorge where it enters into the valley of the North Fork of How- ard’s Creek, a dam 300 feet long and 20 feet high will have to be constructed to stop the w^ater and turn it by a tunnel 200 yards long into the valley of the Middle Fork of Howard’s Creek, after which it will be conducted by a feeder canal 2.8 of a mile long, of a cheap and easy construction, to the summit-level. In the valley of Little Creek, there will also be a reservoir, the water from which will be conducted by a feedt-r of 4.3 miles long into the Anthony Creek reservoir. The mound of this reservoir will be 690 fet-.t long and 40 feet high. The area of the reser- voir will be 127 acres, and the depth of available water 20 feet. Upon the north or main fork of How^ard’s Creek there will be two reservoirs ; the lower one of which is called the Howard’s Creek reservoir, will be made by a mound, 1,180 feet long and 50 feet high. Its area will be 156 acres, and depth of available water 25 feet. The upper one is called the Jericho reservoir, the mound for which will be 222 feet long and 63 feet high, its aiea 92 acres, and mean depth 21 feet ; all the water of which can be drawn off into the lower ri servoir whenever it may be re- quired. The water from these two reservoirs will be conducted by a feeder-canal, 1,800 feet long, into the pond formed by the dam acioss Dry Creek, where it will con- nect With the water from Anthony’s Creek. The filth and last reservoir is that upon Tuckahoe Creek. This reservoir is formed by a mound 589 feet long and 85 feet high; its area is 159 acres, and m» an depth of available water 30 feet. The localities of these reservoirs and feeders will be more fully understood by reference to the accompanying map. Under the mounds of all the reservoirs adequate provision has been made, conform- able with your direction, for culverts which will allovv the waters of the creeks to pass off during the construction of the mounds, and obviate all annojances and danger to the work in time of freshets. After the reservoirs are filled, these culverts, with the ayipendages of pipes and gates, will be used for drawing off the water as required, for the purpose of feeding the canal, or for drawing it all off, if necessary, for repairs. Having given the above brief description of the general jilan of the reservoirs, the next thing lo be considered is the SUPPLY OF WATER. In conformity with the principles laid down by you, I shall first enter into a calcula- tion of the supply of water which will be required for the use of the canal, and then into a calculation of the supply that will be afforded by the reservoirs. The whole length of the canal to be supplied entirely by the reservoirs would be that 756 REPORT OF THE CHIEF OF ENGINEERS. portion between the point on the eastern side of the snminit where Dunlap’s Creek is taken in as a feeder, and the point on the western side of the snrninit where Howard’s Creek is taken in as a feeder, a distance of about 9 miles. But as the supply from Howard’s Creek will be considerably diminished by the quantity of water which will be shut off by the reservoir above, we wdll not consider Ploward’s Creek as affording any supply at all, except what is obtained from its reservoirs. It should, how^ever, be borne in mind that this will be a most liberal deduction from our actual supply, espe- cially during the winter months, when this creek is full, and would be a most important auxiliary. The length of canal then which we wdll assume as supplied by the reser- voirs is all that portion included between the point where Dunlap’s Creek is taken in and the Greenbrier River, a distance of miles. After the prism of the canal shall have been filled, the yearly supply which will be demanded from the reservoirs will be a quantity sufficient to supply the loss by leakage through the locks and evaporation and filtration from the canal, and the quantity consumed in the passage of the boats through the locks of the summit-level. From experiments made by Mr. Fisk on the Chesapeake and Ohio Canal, the loss by leakage through the locks, which are of the same size as ours, amounted to 62 cubic feet per minute, and the monthly loss upon the same canal from eva{)oratiou and filtra- tion was about IPuice the quantity of water contained in it. The w'hole quantity, then, lost upon our canal would be, according to Mr. Fisk’s experiments, 59 cubic feet per minute for each mile. According to Mr. Jervis’s experiments on the Erie Canal, the total loss from evaporation, filtration, and leakage through thegate.s is about 100 cubic feet per minute for each mile. Let us, then, assume the highest of these quantities as our standard. The portion of the canal occupied by the tunnel and its approaches being through solid rock, would be subjected to no leakage, and not as much loss by evaporation as would be supplied by subterranean springs, and is, therefore, excluded from this calculation, leaving the entire length of the canal subject to filtration and evaporation lli% miles. The loss, then, by leakage, filtration, and evaporation would be 1,180 cubic feet per minute, or 1,699,200 cubic feet per diem. In making the estimate for the quantity of water consumed in passing the boats through the locks, let us assume that, the canal shall enjoy a full trade, and the boats pass through the locks at the summit as fast as possible. The average time of a boat passing a lock of 10 feet lift is about 6 minutes, or about 240 boats per diem. Assum- ing a full trade, we must also assume a fair alternation of boats i)assing the summit- level, which wmuld allow 1^ x^nsms of lift to each boat, or 360 locks full of water per diem, which, for locks of 10-feet lift, would amount to 5,400,000 cubic feet per diem. Add to this quantity 1,699,200 cubic feet, the quantity lost by leakage, filtration, and evaporation, and we have 7,099,200 cubic feet, or 262,933 cubic yards, per diem as the quantity of water necessary to navigate the canal with a full trade. To arrive at the amount of water that will be available to supply the above demand, we must estimate the quantity that will be supplied by the rain falling upon the area of country drained by the several streams upon which the reservoirs are to be con- structed. For this purpose rain-gauges have been kei^t at points bordering upon the said streams, and the results of observations of four consecutive years have been recorded. An accurate survey of the area of the country drained was made under my direction during the last summer, in which the whole outline of the basin of each stream extending along the top ridges of the Alleghany and its sxmrs was carefully traced. From the above data the average quantity of rain which falls per annum can be easily obtained ; and after making the prox)er deduction for evax)oration and absorx)- tion, we arrive at the quantity of available water which the streams afford. • But as this supxfiy is not constant, but variable, it is necessary that it should be regulated. For this x:>urpo>e mounds must be constructed at convenient points across these streams, which will dam them up and form large reservoirs, from which the supply requisite for the navigation of the canal can be drawn off’ at x)leasure. To ascertain the quantity of water which these reservoirs will contain, an accurate survey was made of their superficial extent, after which cross-sections of their depth were taken at every con- siderable variation in the ground with the angles of the hill-side at every station of 100 feet. Each reservoir was then divided into a number of fields, the superficial and cubic contents of which were sex^arately calculated. The area of the country drained into these resx^ective streams is as follows: Acres. Into Anthony’s Creek 65, 160 Into Little Creek 5, 634 Into Howard’s Creek 20, 196 Into Tuckahoe Creek 9,522 Total 100, 512 or 157 square miles. The result of the rain-gauges at Anthony’s Creek is an average of 37 inches x^er an- num, and at the White Sulphur Springs 38 inches per annum, giving a general average APPENDIX V. 757 of 37-1^ iiiclies, which, filling upon the above area, will give 506,8 H, 833 cubic yards per annum, or 1,388,627 cubic yards per diem. The auuual drainage from a given area depends upon the climate and the topographical and geological featu -es of the country. From extensive experiments made by Mr. Charles Ellet, jr., upon the Ohio River, which are recorded in his ‘‘ Physical Geography of the Mississippi Valley,” he has ascertained that the annual drainage in that section of country is 40 percent, of the rain thatfalls. In other localities it is found to be 50, and as much as 60 per cent. The country sur- rounding the summit-level of the James River and Kanawha Canal is all mountainous. The sides of the mountains are generally steep and but scantily covered with soil, offering every facility for a rapid discharge of the water which falls upon its surface. When we take into consideration the difference between this country and that which is drained by the Ohio River, the latter consisting principally of gentle slop -s, we are warranted in the conclusion that its annual drainage must be by far the greater of the two, and even greater than any of the above estimates. From observations made by Mr. J. B. Jervis, in reference to the reservoirs for the Chenn.ngo Canal in the State of New York, it appears that, in that locality, al)out two fifths (or 40 per cent.) of toe quantity of rain may be collected for the supply of a reservoir. We will put it, therefore, at the low estimate of 40 per cent., which will leave 555,450 cubic yards per diem as its drainage, and the quantity that may be col- lected in reservoirs, or more than twice as much as would be required for the naviga- tion of the canal. By a proper distribution of r-servoirs nearly the whole of this quantity of water might te collecred and retained for future use. But, as all of it would not be required, it is only proposed to construct a few reservoirs in the most advantageous localities, where narrow passages through the mountains suddenly ex- pand into spacious valleys, and where short ajid high mounds thrown up at no great expense, across the narrow gorges, will shut up large bodies of water, which can be drawn off and used at pleasure. The following table will show the area and cubic contents of these reservoirs, and the annual drainage into each of them : Reservoirs. Area of reservoir ill acres. Cubic yards avail- able water in reser- voirs. Area of drainage in acres. ^ 'S oH.S •2 ^ o Forty per cent, of rain allowed for drainage. Antliony’s Creek 2, 753 127 248 159 109, 189, 130 3, 013, 856 .9, 276, 410 ’7, 693, 464 65, 160 5, 654 6, 75') 9, 522 324, 099, 758 28, 122, 468 33, 573, 870 47, 361,539 129, 039, 903 11, 248, 987 13, 429, 548 18,944,016 Little Creek Howard’s and Jericlio Tuckahoe Total 3, 287 129, 172, 800 87, 086 433, 157, 635 173, 263, 054 From an inspecfion of the above table, it will be seon that the quantity of water assumed to be available, and proposed to be reserved for the use of the canal, is 173,263,054 cubic yards per annum. From this quantity should be deducted the annual loss from evaporation and leakage of the reservoirs and feeders. We will put down the annual evaporation from the reservoirs at Smeatou’s estimate of 36 inches; but 60 per cent, of the annual fall of rain, or 22 inches, having been already allowed to pass off by evajioration and absorption, in which the rain falling upon the surface of the res'-rvoirs is included, only the difference between that <]uautiLy and 36 inches, or 14 inches, should be allowed for evaporation. The allowance for leakage and absoiqition in the reservoirs should be limited to the quantity of leakage through the mounds. There can be no leakage or absorption or filtration in the reservoirs themselves, for after the water has passed through and saturated the rhiu overlying stratum of soil, it would reach the impe'ictrable rock, and then it would have to stop ; there could be no further absorption or filtration. We will then only consider the leakage through the mounds, and for this purpose we will suppose each mound to be the bank of a canal of 30 feet bottom, and that the contents of such a canal would pass through its banks once in fifteen days, eras there would be but one bank, once in a month such an allowance for all the reservoirs would bo equal to 40 inches of their surface annually, which, adiRd to 14 inches, gives 54 inches as the whole deduction for evaporation and filtration in the reservoirs, eipial to 23,863,620 cubic yards per annum. For evaporation and filtration in the feeders, allow that each feeder will lose the whole of its prism of water once in every fifteen days. This for 758 REPORT OF THE CHIEF OF ENGINEERS. miles of feeder, with an area of cross-section of 58.50 square feet, will amount to 2,051,200 cubic yards per annum. We have then for deduction : Cubic yards. For evaporation and filtration in the reservoirs 23,863,020 For evaporation and filtration in feeders 2, 051, 200 Total 25,914,820 Which deducted from total supply 173, 263, 054 Leaves for available water, per annum 147, 348, 234 Or, per diem 40.3, 694 The quantity estimated as required being, x)er diem 202, 933 We have a surplus of water, per diem, over demand, equal to 140, 761 In making the above calculations, we have exact data for the area of drainage and the downfall of rain. The allowances made for the drainage and for losses by evapo- ration and filtration in the canal, reservoirs, and feeders, are based upon careful obser- vations and experiments made by distinguished engineers in this country and in England. In adopting this basis, however, it will be observed that we have not availed ourselves of the highest or lowest estimates which would best suit our pur- pose, or even of an average, but have been contented to assume the lowest for the amount of drainage, and the highest for the evaporation, filtration, and leakage. Still we have the large surplus of 140,761 cubic yards per diem, or more than one-half of the required supply. This surplus is amply sufficient to cover any contingencies or objections that ingenuity may suggest. * To cover the ground, however, more fully, it is a fact worthy of attention that we have not in the above estimate a.vailed ourselves of the whole area of drainage, but that there still remains 13,446 acres of drainage on the middle fork of Howard’s Creek, which might be added to the above estimate, and that by multiplying the reservoirs in the valleys of the streams embraced in the sur- vey, a sufficient supply of water could be obtained for nearly two such canals. If, at some future time, the increase of trade on the canal should demand a double set of locks, and consequently nearly a double supply of water, the additional quantity that might be demanded could be obtained in the mode above indicated. Dunlap’s and Potts’s Creeks on the eastern side of the mountain could also be taken in, if necessary. Below is an estimate of the cost of supplying the summit-level : Anthony’s Creek mound $166,181 32 Anthony’s Creek feeder-tunnel 318,509 20 Little Creek mound 17, 056 25 Little Creek feeder 43,996 10 Jericho mound 15,030 00 How^ard’s Creek mound - 47,554 50 Howard’s Creek feeder 5,799 20 Dry Creek mound 2,435 00 Feeder from Dry Creek to Summit, including tunnel 52, 382 00 Tuckahoe mound 83,348 50 Land-damages 51,000 00 803,292 07 Add for contingencies 20 per cent 160, 658 41 Total cost 963,950 48 The above estimate includes all the reservoirs and feeders that may be necessary for the navigation of the canal with a full trade, and is intended to cover all contingencies. But as the canal at first would not have a full trade, and would probably not be open on an average for more than eleven months in the year, thereby diminishing the re- quired su])ply of water one-twelfth, it is evident that there would be no necessity for constructing all these reservoirs at first. Anthony’s Creek reservoir would of itself be amply sufficient to supply the canal during the whole year — for many years to come. A strict estimate of the cost of supplying the summit-level should then be confined to The cost of that reservoir and its feeders, which will amount (with 20 per cent, added for contingencies) to $696,963. Daily gauges of Anthony’s Creek were commenced last August, and will be continued until twelve mouths shall have expired. These gauges, together with the rain-gauges, will furnish us with exact data, by which we can ascertain the annual discharge of water by the creek, the difference between which and the downfall of rain will show'' the quantity carried off by evaporation and absorption. Toward the close of the survey I was joined by Professor Tuomey, of the University of Alabama, who was invited by the President to make a geological examination of APPENDIX V. 759 the sites of the reservoirs. That gentleman made a laborious and careful examination of the whole ground, the results of which have been published, and are highly satis- factory. Very respectfully, your obedient servant^ E. Lorraine, Principal Assistant Engineer. Col. Walter Gwynn, Chief Engineer James Bi ver and Kanawha Canal. Approved and submitted in lieu of any report of my own, as is usual upon surveys made under my direction. Walter Gwynn, Chief Engineer James Biver and Kanawha Canal. Table of rain-gauges. At Anthonj^’s Creek. At White Sulphur Springs. 1847- ’48. 1848-’49. 1849- ’50. 1850- '51. 1847- ’48. 1848- ’49. 1849-’50. 1850- ’51. Meters. Parts. 1 Meters. Parts. Meters. Parts. Meters. Parts. Meters. Parts. 'o Parts. Meters. Parts. Meters. Parts. September .... 22 6 24 5 11 3 6 2 21 5 20 4 16 0 6 2 October 3.3 0 11 3 34 3 21 5 30 7 20 7 30 3 21 5 November 44 4 35 4 11 0 31 6 42 7 35 3 10 0 31 6 December 32 1 42 1 56 6 46 3 37 8 43 5 56 0 44 2 .J.anuary 23 0 25 5 38 7 5 0 22 7 26 5 36 3 3 5 February 33 3 23 7 31 9 32 2 32 2 22 5 26 8 32 9 March 44 6 42 9 23 0 24 4 48 2 40 8 23 3 24 4 April 17 7 19 3 35 8 28 4 21 6 20 8 45 6 28 2 May 51 9 32 3 53 1 40 5 39 9 33 1 53 1 27 2 June 11 9 20 5 40 1 17 7 16 0 39 0 25 6 25 5 July 41 7 58 6 12 0 49 6 57 2 45 8 30 6 86 0 August 38 7 35 0 42 0 18 8 '36 2 13 2 46 1 18 8 Total 394 9 371 1 388 8 322 2 406 7 361 6 399 7 350 0 Inch 39. 49 37. 11 33. 88 32. 22 40. 67 36. 16 39. 97 35. 00 SUPPLEAIENTAL report on the survey of the summit-level, by MR. E. LORRAINE, ASSISTANT ENGINEER JAMES RIVER AND KANAWHA CANAL. Pattonsburgh, October 5, 1852. Sir : I had the honor of submitting to you in January last a report upon the survey of the summit-level, and upon the supply of water which might be obtained for the nav- igation of the canal. The calculations for the water-supply were based upon accurate surveys of the basins of the various creeks from which it was supposed to feed, and upon the observations of the downfall of rain for the four preceding years. To the engineer or to the man of science this method is iierfectly satisfactory, be- cause he knows it is customary, and that where it has been tested the practical results have always proved that the theory was correct. But to the person unaccustomed to scientific investigations the whole process appears to be chimerical, at least, if not em- inently ridiculous. Such persons regard the scheme of carrying the canal over the Alleghany Mountains more as one of the reveries of a lunatic than as the sober thought of a sound practical mind. And the notion of its impracticability is based mostly upon the idea that it will be impossible to obtain water sufficient to supply the summit-level. The enemies of the canal have taken advantage of the ignorance of the i)eo)de, and en- deavored to plunge them into still greater darkness, eith^er by the misrepre^ntation of facts or by the suppression of the truth. The country about the summit-level has been represented as a region of parched and arid rocks, upon which the rain never falls, or if it does, only to be swallowed up by innumerable chasms. The creeks and streams are said to be all (fc-y, and the only hope of supplying the canal is from a few holes to be dug in the mountains, which are to be filled from the mists and dews of heaven. The object of the survey which you caused to be made last year was to remove all doubt as to the practicability of prosecuting the water-line to the Ohio. A report was made upon that survey, which I believe satisfied yourself and every candid friend of truth that the water line was practicable, and that the summit-level could be amply supplied at a very moderate cost. The estimates, however, of the supply of water were based upon the downfall of rain. But as it has seemed heretofore a great mys- 760 EEPOET OF THE CHIEF OF ENGINEEES. tery to the people of Virginia how it was possible" for a canal to be supplied by rain, forgetting that the Mississippi River receives its annual supply from the same con- temptible source, it was desirable that the subject should be investigated in another way, which should be more practicable and tangible to the eyes of even those who do not wish to see ; and as there seems to be such an objection to the canal being sup- plied in the same way as our rivers, lakes, and seas, it was hoped that if it could be clearly proved that Anthony’s Creek itself, apart from all connection with the clouds and rain, actually discharged enough water to supply the canal, that then the mystery would be solved and the most incredulous would become convinced. For this purpose you instituted daily gauges of Anthony’s Creek, to be continued for one year. Accu- rate sections of the creek were taken in three different places, and it was so arranged that it was only necessary that the width and velocity of the creek should be measured daily and reported by a careful and intelligent person. Such a person we found in Col. Andrew Humphreys, who deserves great praise for the punctuality and faithful- ness with which he has discharged the important duties which were assigned to him. The gauges have been regularly^ taken, the reports have been sent in, and are now before me with the calculations deduced from them, which I now beg leave to submit to you. Table of the qmntitij of neater discharged by Jntliony’s Creelc in one year. January. .. P'ebruary.. March April May June July August September. October . . November. December. Months. Cubic yards of water dis- charged per It, 649, 673 40, 628, 408 38, 455, 285 45, 333, 023 13, 262, 9.39 19, 208, 005 4, 586, 482 7, 071,220 1, 194, 709 780,491 6, 963, 657 21,393, 063 Total, 210, 526, 955 The above is the quantity of water which actually flowed down Anthony’s Creek in one year. In making my calculations of the supply of water from the downfall of rain, I as- sume an estimate of 40 per cent, as the quantity of rain that would be drained off into the creeks and that could be collected in reservoirs. I assume this quantity because it is the same as was reported by Mr. Ellet as the drainage of the valley of the Ohio, and by Mr. Jervis as the drainage into the reservoir of the Chenango Canal. I, how- ever, stated that the drainage in other localities has been found to be as much as 60 per cent, of the rain that fell, and ventured the assertion, from ray knowledge of the lieculiar formation of the basin of Anthony’s Creek, that its drainage would amount to even more than 60 per cent. In that conclusion I am now sustained by the facts which have since been elicited by the gauges of the creek. The rain-gauges for the same year give 34.23 inches as the downfall of the rain, equal to about .300,000,000 cubic yards of rain, of which about 210, .500, 000 cubic yards, or 70 per cent., was drained off and dis- charged by the creek. I observe, however, upon a comparison of the rain-register and the gauges of the creek, that the latter was sometimes very much swollen from rains not indicated by the rain-register, and which must, therefore, have fallen near the head- waters of the creek. That being the case, it is safest to assume the downfall of rain at what it has averaged for the last five years, viz, 36.366 inches. It is also proper to deduct froi^i the annual discharge of the creek a constant quantity equal to 365 times its least daily discharge, which is about 2,000,000 cubic yards per annum, which may be considered as supplied by springs. This quantity being deducted, leave s the drain- age into the creek equal to 208,526,955 cubic yards. The downfall of 36.386 inches gives 318,738,394 cubic yards of rain. The drainage is therefore 65^ per cent, of the downfall of rain. This is a mere matter of philosophical inquiry, in no way connected with the supply of the summit-level, as it is intended in the present investigation to ignore rain altogether until it becomes creek-water. It is onl}" alluded to as establishing an in- teresting meteorological fact, based on carefully-conducted experiments on a very large scale. To those persons who have heretofore been impressed with the idea that the streams about the summit level are nearly dry or lose themselves in immense chasms, it may APPENDIX V. 761 appear strange that so great a quantity of water should pass down Anthony’s Creek. To those who live upon the creek it will he nothing new. They have seen it as I have, when swollen by rains, rushing through the “Narrows” with an impetuosity and volume almost incredible. I have myself seen it discharge in one day a sufficient quantity of water to till the entire canal from Buchanan to Richmond, or 200 miles of canal 40 feet wide and 5 feet deep. I have also seen it wide enough, deep enough, and swift enough to carry seven of our largest freight-boats abreast at the rate of 5^ miles an hour. The quantity of rain which fell last year was only 34^ inches, which is considerably below the average. Neither was there any remarkable freshet in the creek beyond that in common to it every year. We may therefore assume the above quantity of 210^ millions cubic yards are the average yearly supply. The question then is, is it enough for the wants of the canal ? I have estimated the quantity of water necessary for the supply of the summit-level at 9.5,970,545 cubic yards per annum ; this was supposing a boat passed the summit- level every six minutes in every day for 365 days, which is a very extravagant calcu- lation ; for if you allowed every boat to carry 50 tons it would make an annual tonnage of 4,380,000 tons. Cubic yards. The quantity allowed for filtration and evaporation in the five reservoirs and feeders was 25,914,820 Which added to 95, 970, 545 Gives 121, 885, .365 as the total quantity to he supplied. Anthony’s Creek yields us 210, 526, 955 We therefore have a surplus of 88, 641, .590 for contingencies. In the above calculation I have allowed for filtration and evaporation in five reser- voirs, which is also ah extravagant amount, as we are supposing but one to be in use. It is an undeniable fact, then, that Anthony’s Creek alone affords a sufficient quan- tity of water to supply the summit-level. If that were the only stream upon which we had to depend there wmuld be no obstacle to the passage of the canal over the mount- ains. But there are three other creeks, viz. Little Creek, Tuckahoe and Howard’s Creeks, whose united volume amounts to about one-third of Anthony’s Creek, which, if necessary, could be appropriated for the use of the canal. Anthony’s Creek being proved sufficient, the only question, then, is, can it be made available at a reasonable cost? If a Virginia farmer has a small stream running through his plantation, and he wishes to erect a mill upon it, all that he does is to build a dam across the valley of the stream and put his mill up near it, nothing doubting that the dam will stop up the water and form a pond, from which the desired water-power can be obtained. He does not em- ploy a geologist to tell him whether his pond will hold water or not. He goes ahead, with perfect confidence that the water can be dammed up and made available. TUe James River and Kanawha Company can go ahead with the same confidence and dam up Anthony’s Creek. The operation is exactly the same, only upon a large scale, and with the additional security that the basin of the creek has been examined by an emi- nent geologist, who pronounces the rocks to be in the best possible position to retain the water. After the mound is niade, the craek itself will fill the reservoir, and keep it full, besides supplying all the demands of the canal. It will require a tunnel 2^ miles long to conduct the water from the reservoir to the canal. This tunnel is through slate rock, and need only be 6 feet in diameter. The tunneling-machine lately put into operation at the Hoosac tunnel would walk through it in a very short time, or, even if it had to be excavated in the usual way, it could be done at a cost of about $300,000. The whole cost of supplying the summit-level wil not exceed $700,000. All the fancied difficulties and impossibilities of supplying the summit-level, and consequently of carrying the canal across the mountains, vanish into thin air before this array of facts. It becomes, then, a mere question of policy or expense. In that light the subject ceases to come within the legitimate scope of this report. As you thought proper to assign me the duty of making the survey, I considered it within my province not only to submit the above facts, but the deductions to be drawn from them, and to dispel, as far as possible, the thick medium of prejudice and ignorance through which the subject has been heretofore viewed, not only by the enemies of. the canal, but, I believe, by some of its best friends. Very respectfully, your obedient servant, E Lorraine, Assistant Engineer J. li. <^- K. Canal. Col. Walter Gwynn, Chief Engineer J. B. cf K. Canal. 762 REPORT OF THE CHIEF OF ENGINEERS. LETTER FROM THE SECRETARY OF WAR, TRANSMITTING A REPORT ON THE JAMES RIVER AND KANAWHA CANAL-ROUTE. Doc. No. 216. Ho. of Reps.; War Department; 20th Cong., 1st Sess. March 24, 1828. — Read and laid upon the table. Washington : Printed by Gales Seaton, 1828. War Department, March 24, 1828. Sir: In compliance with a resolution of the House of Representatives, of the 2d of January last, I have the honor to transmit herewith a letter from the Chief Engineer, of this date, accompanied by a report on the James and Kanawha canal-route. To save time, it has been thought necessary to transmit the original report, which report, it is requested, may be returned to this Department, when it shall have answered the purposes of the House. I have the honor to be, very respectfully, sir, your most obedient servant, James Barbour. The Hon. Andrew Stevenson, Si)ealcer of the Rouse of Representatives. Engineer Department, Washington City, March 24, 18,8. Sir : I have the honor to lay before you a letter from Capt. W. G. McNeill, of the topographical engineers, accompanied by his report on the route for a canal between the James and Kanawha Rivers, which report was called for by a resolution of the House of Representatives of the 21 Januaiy last. I have the honor to be, very respectfully, your obedient servant, Alex. Macomb. Major-General, Chief Engineer. Hon. James Barbour. Secretary of War. Georgetoavn, D. C., March 24, 1828. Sir: I have the honor to transmit herewith a report on the James and Kanawha Canal, in illusTration of which a map is now in the course of preparation. All the maps and profiles relating to the experimental surveys are already completed, as well as those exhibiting the locations from the James to the Greenbrier River; the others, including the locations down the Greenbrier and New Rivers, are in progress of completion. There are in all nearly thirty large maps, including the plans and profiles, on a scale suitable to the exhibition of details; but they are too unwieldy to accom- pany the report, and are retained till they can be reduced into one general map. Most of the drawings in relation to the Roanoke and Kanawha Canal are nea ly finished, but their size renders it necessary that they, also, be reduced into a more convenient shape. I have the honor to be, sir, with great respect, your obedient servant, Wm. G. McNeill, Captain United States Topographical Engineers. Maj. Gen. Macomb, Chief Engineer, Washington City. To Major-General Macomb, Chief Engineer: Sir: In pursuance of instructions from the board of internal improvements, of July 29, 1826, predicated on the orders of the Chief Engineer, assigning to my brigade the execution of certain surveys, “ to ascertain the practicabdity and means of connecting the waters of the Great Kanawha with those of the James or Roanoke River, by canals or railroads, or both, and also the connection of the Roanoke and James,’' I have now the honor to REPORT that the short interval which elapsed between the receipt of my instructions and ter- mination of the first season was consumed in preliminary operations, which merely related to the first object to which my instructions referred, to wit, the practicability of a “ connection of the Kanawha and James Rivers, by means of a canal;” that in APPENDIX V. 763 May, 1827, the surveys were resumed and continued till November ; and that in that period the required location of a line of canal from Covington, on James River, to the foot of the Great Falls of Kanawha was effected, with that of all the works relating to a summit-level, such as feeders, reservoirs, &c., and the consequent location made, as enjoined, in the event of its practicability, of a canal from the forks of the Roanoke to the Great Falls of Kanawha River. The subject under consideration naturally advances as our first inquiry, “theprac- UcahiUty of a connection of the Kanawha and James Rivers, by means of a canal, and, in contemplating such a connection, we are led, in the first place, to remark that the characters of the small tributaries which empty into the James or Greenbrier Rivers, (the country intermediate to those rivers being that under consideration,) like those of most mountain-sti’eams, incontrovertibly prove their inadequacy to the per- manent supply of a canal, and that, in consequence, the practicability of the project must depend on the adequacy of the Greenbrier River for that purpose, at some point whence ils waters may be conducted to a summit-level of such elevation that the dividing-ridge may be passed without encountering an excessive length of tunnel. To the solution of this question, then, various experimental lines were directed through the ravines, which, on the one side, bound the tributaries of Dunlap’s Creek, and, on the other, through those which, heading opposite to- them, define the course of eit her Second Creek, Howard’s Creek, or Anthony’s Creek, in their progress to the Green- brier River, by which was ascertained the height of the dividing ridge in its greatest depressions, and a profile was carried up Greenbrier River to obtain such knowledge of its character as was indispensable to the judicious selection of a summit-level. A comparison of the elevation of the dividing ridge, on any route examined, with that of the Greenbrier River, at any point within a reasonable distance, at once dis- closes the necessity of a tunnel ; and that this, with every other fact connected with the subject, may be known, I proceed to detail the results developed in the course of our investigations. As the basis of comparison in the description about to be given, all the heights and distances will be refer) ed to the bench-mark, at the mouth of Dunla|)’s Creek, oppo- site to the town of Covington and to begin with that route which, from the near approach of the opposite streams, the gradual slope on either side of the dividing ridge, and the seemingly great depression of the Alleghany Mountains, might induce a belief of its paramount advantages, we shall describe — FIRST — A ROUTE BY THE VALLEYS OF DUNLAP’s AND SECOND CREEKS. Dunlap’s Creek, or its main branch, (known as its south fork, on which are the famed Sweet Springs,) heads nearly on the summit of the ridge which divides it from Second Creek, and pursues its course at the base of Peter’s Mountaiu, nearly parallel with the Alleghany Mountain, till within 6 miles of its mouth, when, having received the waters of Ogley’s Creek, it continues in a direction in general nearly at right angles to its former course. Between its source and its conffuence with Ogley’s Creek several small runs contribute to its supply, such as Cove Creek, Fork Run, and Brush Creek ; but they are noticed rather on account of their directions, which shall hereafter be alluded to, than because of any efficient aid to be derived from them toward the feeding of a canal. Dunlap’s Creek is, in general, bordered by flats varying in width from 2r)0 yards to half a mile, although it may sometimes occur that the hills impinge so far upon the stream as to render it preferable to gain an opposite flat by an aqueduct (since bottom- land is always to be found on one side or the other, and the width of the stream never exceeds 35 yards) than to encounter the obstacles presented by steep side-lying ground in an attem})t to avoid that expense. This remark, however, is applicable in a degree as well to all the other valleys through which experimental lines were run as to that of Dunlap’s Creek. For the first 22 miles the average rise in the stream may be assumed at 25 feet per mile, whe!) for the next mile the regularity of its ascent is interrupted by falls and rapids comprising 135 feet ; thence to the summit succeeds, first, an ascent for 8 miles of 73 feet per mile, and, in the remaining half a mile, a rise of 105 feet; making the distance of the summit from the base-mark 31^ miles, and its elevation 1,372 feet 7 inches. A depression in the Alleghany Mountain west of that just alluded to induced an- other experimental line from Dunlap’s to Second Creek ; but of this suffice it to say it was found still higher, being 1,408 feet above the base-mark. * The bench-mark at the mouth of Dunlap’s Creek is 12 feet below a bench-mark established on the opposite side of the river, in 1819, by Messrs. Moore and Briggs, who reported the mouth of Dunlap’s Creek to be 1,238 feet above tide-water. Our base- mark, therefore, which is about 2 feet above the water, may be assumed to be 1,240 feet above tide- water. 764 REPORT OF THE CHtEF OF ENGINEERS. On tlescendiag Second Creek tlie slope from the summit to the Greenbrier River av- erages 337 feet per mile, with the exception of the first two and a half miles, in which the fall is 140 feet, or, in otlier words, the distance from the summit of the mount- ains to the mouth of Second Creek is 27 miles and the fall 067 feet, making the total distance from the mouth of Second Creek to that of Dunlap’s Creek, by their re- spective valleys, fifty-eight and a half miles, and the Greenbrier River,* at the mouth of Second Creek, 406 feet above the base-mark. Our present i)urpo8e, however, being merely, in the first place to show the relative altitudes of those depressions which suggested the different experimental lines across the mountain, and to compare them with the elevation of the Greenbrier River, that we may assume a summit-level which may command the waters of the Greenbrier without involving an impracticable length of tunnel, we shall confine ourselves to the few facts already stated, till, in succession, each route shall have as briefly been con- sidered. For obvious reasons, which will appear hereafter, (see page 767,) no other experi- mental lines were run from Dunlap’s to Second Creek, nor was it deemed material, after a careful reconnaissance, that any tributary above Fork Run should be surveyed. We shall then next describe that as the SECOND ROUTE — BY DUNLAP’s CREEK, FORK RUN, AND HOWARD’S CREEK. Howard’s Creek, the first tributary of any note to the Greenbrier River above Sec- ond Creek, at the distance of seven miles from its mouth, is formed by the union of its three branches, which, from their relative directions, we shall designate as the South, the Middle, and the North Forks. The South Fork rises near the heads of the upper branches of Dunlap’s and Second Creeks, and pursues a course nearly parallel with Dunlap’s Creek till it receives the wafers of Tuckahoe Run, a small stream which empties into it near Comb’s saw-mill. The first depression remarkable in the dividing-ridge between Dunlap’s Creek and the South Fork of Howard’s Creek is found where the Alleghany is indented on the west by Tuckahoe, as it is on the east by Fork Run, an opposite •tributary to Dunlap’s Creek ; and it is through this depression that the second route is directed. That part of the route between the mouth of Dunlap’s Creek and Fork Run has already been described as ascending at the average rate of 25 feet per mile ; thence to the summit of the mount- ain through the narrow valley of Fork Run (sufficiently wide, however, for a canal, the flats usually being from 100 to 200 feet wide) we ascend at a much more rapid rate, and, in the short distance of five miles and thirty-eight yards, rise 689 feet, mak- ing the height of the Alleghany, at the head of Fork Run, 1,092 feet, and its distance from the base-mark twenty-two and three-quarters miles. The descent on the west, by the ravine of Tuckahoe Run, is very precipitous for 1 mile 593 yards, the fall, in that distance, being 331 feet ; when, having arrived at the South Fork of Howard’s Creek, the fall becomes quite gradual; the distance from the mouth of Tuckahoe to that of Howard’s Creek being lOf miles, and the fall 315 feet. The valley of Howard’s Creek is, in general, wide, and under cultivation, and, with the exception of the short distance of one-eighth of a mile above Hunter’s Mill, where the mountains on either side terminate in the stream, the construction of a canal would encounter no particular difficulty from the nature of the ground. Fork Run, at the distance of 4 miles 750 yards from its mouth, branches in two direc- tions, that which we have heretofore alluded to being the soutliern of the two; but as the Alleghany, where it divides the northern branch from the Middle Fork of Howard’s Creek, presents another depression, through it was directed the THIRD ROUTE — BY DUNLAP’S CREEK, FORK RUN, TO THE MIDDLE FORK OF HOWARD’S CREEK, AND THENCE BY THE VALLEY OF H;>WARD’S CHEEK TO THE GREENBRIER RIVER. Of this route, although among the most promising, since, wirh the exception of a few miles, it occupies the same ground as the second route, but litcle need be said. The height of the dividing-ridge is greater than by the second, being 1,221 feet, and the access to the summit, on either side, more gradual. It does not, therefore, present as short a tunnel. Indeed, the only advantage which it possesses arises from its being nearer the Greenbrier River; and it will be shown hereafter how far that can place it in competition. Brush Creek, the next tributary to Dunlap’s Creek, has been enumerated among those whose direction claims some notice. But since the turnpike road, which pursues the valley of the South Fork of Ogley’s Creek to its source in the ridge, crosses that ridge and enters the valley of Brush Creek several miles from the summit of the Alleghany, * Viz, the bench-mark at the mouth of Second Creek. APPENDIX V. 765 instead of running a profile from the month of Brush Creek, it was thought sufficient to continue the prodle from Ogley's Creek in the direction of the turnpike, and from the ])oint at which it strikes Brush Creek to cross the Alleghany in the three deyires- sions in which the several branches of Brush Creek head oyiposite the Middle and the North Fork of Howard’s Creek ; for it was apyiarent, if a tunnel should be requisite hy either of the routes from Bri;sh Creek, that its length, unless immoderate, Avould be included betAA^een the intersection of the turnpike with that stream and its correspond- ing elevation west of the mountain. The first line from Brush Creek was ditected through the gap occupied by the pres- ent turnpike road to the Middle Fork of Howard’s Creek. It determined the height of that gap to be 1,288 feet. By the second line, the height of the depression through which the former road was located AvaS found to be 1,252 feet ; and, by the third line from Brush Creek, the elevation of the Alleghany, in a deyiression betAveen it ami the North Fork of HoAvard’s Creek, AA’as found to be 1,534 feet. It will be seen, on the a‘-sumpti()n of a summit-level, that by no route from Brush Creek could the passage of the Alleghany be effected by as shoi t a tunnel as Avill be found on other routes. Ogley’s Creek, the last tributary to Dunlap’s Creek of any importance, alone remains to be spoken of. We have already incidentally mentioned, in describing the routes by Brush Creek, that its southern fork heads in a ridge between it and Brush Creek ; but although the height of that ridge (it being 1,143 feet above the base-mark and 200 feet above its western base) Avould, of course, necessitate a longer tunnel from the South Fork of Ogley’s Creek than from Brush Creek, and would, in consequence, opyiose itself successfully to the location of a canal, yet, since the shortest route from the James to the Greenbrier River would be in the direction of the present turnpike road, in refer- ence to the selection of the best route for a railroad, (an ulterior object contemplated by my instructions,) the valley of the South Fork of Ogley’s Creek merits more consid- eration than if a canal alone were the subject of our investigations. The valley of Ogley’s Creek is not, either in quality of its soil or the nature of the rocks in its vicinity, particularly distinguished by any peculiar characteristics; tor 3 miles from its mouth it affords a considerable quantity of rich productive laud on the northern side, to AA^hich the flats a.(e almost exclusively conflned ; and beyond that, as Ave approach the mountain by its South Fork, we find the valley contracted to a width of but from 50 to 100 yards. The Middle Fork of Ogley’s Creek, however, is the main branch, and it presents a wnder A'alley throughout its course than that of the South Fork, with which it unites 9 miles from the base-mark, and 230 feet above it. Its direction corresponds nearly with that of the North Fork of Howard’s Creek, and is such as to bring it nearer to Anthony’s Creek than any other of the eastern tributaries ; which fact rendered its examination the more important because of the advantages which would result from the proximity of a summit-level on Anthony’s Creek to the source whence the supply of Avater is to be derived. A profile AA^as, therefore, continued throughout the whole length of the Middle Fork of Ogley’s Creek, and thence down Anthony’s Creek, Avith the intent to learn, as well if any depression existed between Ogley’s and the North Fork of HoAvard’s Creek, as the difficulties opposed to the route of a canal in the valley of Anthony’s Creek. Of the one or two experiments made from the Middle Fork of Ogley’s Creek to the North Fork of HoAvard’s Creek, it is sufficient to state that although carried no great distance from the former, they conclusively proved that the height and Avidth of the mountain betAveen those streams rendered their connection impracticable. The eleA^ation of the Alleghany, where it divides the Middle Fork of Ogley’s from the nearest head branch of Anthony’s Creek, is 1,772 feet, and the distance to that point from the base-mark 20 miles and 1,100 yards; the descent thence to the base-mark at the mouth of Anthony’s Creek (5 feet above the stream) was found to be 1,212 feet, and the distance 18-^ miles ; hence the Gret ubiier River at the mouth of Anthony’s Creek is 555 feet above the base mark. The profile from Ogley’s Creek to the mouth of Anthony’s Creek descended for 1^ miles a mere ravine, Avhich bounds the small tributary known as Laurel Run ; but thence to AA ithin 6 miles of the mouth of Anthony’s Creek, during Avhich, from the accession from tributaries, its supply has become considerable, Ave in general find the stream bordered by fertile bottoms, under cultivation, from one-quai ter to a half mile in width. Nor is it until it breaks through the Greenbrier Mountain (when the flats disappear) that either its descent is so rapid or its valley so contracted as to qualify it advantagiiously for the purposes of a reservoir. It is only necessary noAv to add that a profile Avas carried up the North Fork of Ogley’s Creek, till having risen 841 feet above the base-mark, we Avere satisfied of the iuutility of proceeding farther, (since its source was knoAvn to be as elevated and more distant from Anthony’s Creek than that of Middle Fork,) and it will be apparent, from an in- spection of the map that no depression can exist in the ridge dividing the Avaters of Dunlap’s Creek from those Avhich flow into the Greenbrier River, Avhich has not been sufficiently examined to enable us to determine the best route for the passage of the Alleghany — abstracting such considerations as relate to the supply of water. 766 REPORT OF THE CHIEF OF ENGINEERS. These considerations now sn<;gest as the subject of our next inquiry the character of the Greenbrier River, whose relative position to the different routes, or comparative supply at different elevations, is to be so influential in determining the proper eleva- tion to the summit level. The survey and level of Greenbrier River were commenced at a base-mark at the mouth of Howard’s Creek, (5,9.35 feet above the then surface of the water,) and con- tinued for 39 miles up the river, in which distance the ascent was found to be 318 feet, or the point at which we stopped is 770 feet above the base- mark. Anthony’s Creek and Spring Creek are the only tributaries of mucli consequence intermediate to Howard’s Creek and the “ Droop Mountain,” near which latter point the survey of Greenbrier River was discontinued, and, therefore, it was with less surprise we observed that the supply of water did not diminish very materially in our progress up the river, or that, on the contrary, it seemed to vary but little when, in passing over ledges, its whole volume could be estimated. For, as an exception to the apparent uniformity of its supply, it is to be remarked that, at times, the stream almost disappears amojig the loose stones and fissures of its bed, and that, in one case where a vein of limestone traverses its valley, all the water is passed through a subterraneous chanuel for the distance of a mile nearly. This, however, occurs but once, (in the twenty-third mile from Howard’s Creek,) and then only in very low stages of the water. Greenbrier River pursues a very sinuous course through a valley unusually con- tracted, when viewed with reference to the size of the stream, and, with some partial exceptions only, we find it, throughout the extent at present under consideration, bor- dered by high and rugged hills which descend steeply to the water’s edge. There are, it is true, narrow strips of alluvial on one or the other side, but they are never contin- uous for any considerable distance, nor of such value as t > merit consideration when their submersion shall be the consequence of such dams as might be desired for the formation of reservoirs. The country west of Greenbrier River, although hilly, affords rich and arable lands ; east of the river it is mountainous, and in general but illy-adapted to agricultural purposes. But, for a more minute descriirtion of the quality of the soil, vegetable and mineral productions, &c., and of the character of the country above Droop Mountain, resort may be had to the report of Lieutenant Dillahunty, which is appended to this report. A reference to the profile of the river will show that the general and average rise of 7,6 feet per mile for 31 miles above Howard’s Creek is but seldom interrupted ; and that beyond that, or as we approach Droop Mountain, the average rise of the Greenbrier is about 10 feet per mile. The heights to which freshets rise above the ordinary bed of tie stream vary, of course, with the slope and ivklth of the valley. Traces of such as are known to be of frequent occurrence, at almost all seasons of the year, were perceived 6 and 8 feet above low-water mark, and indications were re- marked of freshets having sometimes attained the height of 16 feet in the broadest parts of the river. The influence of but comparatively slight rains is very perceptible in the floods which succeed; and from their freqirency and the magnitude of the volume which passes at such times, we have ample assurance that the most extensive reservoirs (for the formation of which the valley of Greenbrier is so admirably adai^ted) could be replenished as often as might be necessary. To revert, however, to the relative elevation of the Greenbrier River, and of the Alleghany Mountains, in the different depressions heretofore recited, we readily per- ceive that the supply from Greenbrier, on which, as was premised, the practicability of the project rests, cannot be commanded within a reasonable distance without resorting to a tunnel; for the greatest depression in the Alleghany Mountains being 1,092.75 feet above the base-mark, while the elevation of Greenbrier River 39 miles above Howard’s Creek is but 770.7 feet, at an average rise of 10 feet in the Greenbrier, beyond the point to which it was leveled, an elevation corresponding to that of the most clepres-ed point of the Alleghany would not be attained in less than 71 miles from the mouth of Howard’s Greek, and supposing thespurickich projects from the Alleghany, between the waters of Howard’s and Anthony’s creeks, (p. 768,) sometimes called the “ Greenbrier Mountain,” (see map of experimental surveys,) to be passed without winding around its extremity near the mouth of Howard’s Creek, the length of a feeder from the summit-level could not be much, if any, less than would be the distance from the mouth of Howard’s Creek to the assumed elevation of the summit-level on the Greenbrier. Besides, the ruggeduess and steepness of the hills bordering the Greenbrier may be said to be almost in proportion to the elevation above the stream at which we encoun- tered them; and on that account it would be desirable to run the feeder as low as pos- sible, adding to these considerations the fears which might reasonabW be entertained of the inadequacy of the Greenbrier to supply the losses of so long a feeder as would be required to pass the Alleghany without a tunnel, and the necessity of a tunnel would seem to be obvious. This conclusion leads us to compare the lengths of tunnels at different elevations; APPENDIX T. 767 ■which, limiting the depth of cutting at which the tunnels are supposed to commence and terminate at 35 feet, are now exhibited in the following table : TABLE. 6 c Vx O u Designation of route. s| o > r= O ■-M .2 ® = c; — as > 'Z ^ i I I Length of out east of ; the Alleghany Moun- l tains. 4-1 o a C/ S V. c tx. Z the Alleghany Moun- 1 tains. Length of tunnel. Feet. Miles Yards. Miles. Yards. Miles. Yards. 1 Erom Dunlap’s Creek to Second Creek 750 1 283 0 1, 173 10 867 2 From Dunlap’s Creek by Fork Kuii to south fork of Howard’s Creek 750 0 400 0 547 2 210 3 From Dunlap’s Creek by Fork Run to mid- dle fork of Howard’s Creek 750 0 400 0 1,000 3 800 4 From Dunlap’s Creek by middle fork of Ogley’s Creek to Anthony’s Creek 750 0 727 2 943 4 883 5 From Dunlap’s Creek by Ogley’s to the north fork of Howard’s Creek.. 750 0 383 0 1,285 5 1,523 6 From Dunlap’s Creek by Ogley’s Creek to the middle fork of Howard’s Creek 750 0 383 0 1,000 6 783 1 700 1 283 0 1,217 11 2 700 0 430 0 1,017 3 333 3 700 0 430 0 901 4 317 4 700 0 717 3 500 5 1,716 5 700 0 300 0 860 6 1, 187 6 700 0 300 0 901 6 1, 500 1 650 0 440 0 1,550 11 933 2 6.50 0 350 1 73 4 3 650 0 350 0 1,013 5 1.50 4 650 0 684 1 793 9 773 5 650 0 30 0 1,320 7 760 f) 650 0 30 0 1,013 7 547 1 600 0 293 0 1, 650 11 1,277 2 600 0 407 1 964 5 600 3 600 0 407 1 190 5 1,217 4 600 0 660 1 427 10 1, 067 5 600 0 283 1 156 8 720 6 600 0 283 1 190 8 383 By reference to the foregoing table, it is at once perceived that within the assumed elevations of a summit-level a shorter tunnel would be requisite to pass the Alleghany Mountain by the second or third routes than by either of the others ; and it will also be seen that the length of tunnel increases very rapidly on either of those routes as we descend below 700 feet, while at any greater elevation the diminished length of tunnel does not constitute a sufficient advantage to compensate for the increased lockage, the greater length of feeder, or to induce us to forego the advantages of a reservoir on the Greenbrier in its passage through the Droop Mountain, when the fall of the river is known to be greatest, and its valley so contracted that the least area in proportion to the contents of the reservoir would be exposed to the influence of evaporation. We therefore assumed a summit-level at the elevation of 700 feet above the base- mark, and from its western end on the second route a line of feeder was traced at an inclination of 6 inches per mile to its intersection with the Greenbrier River. This line from the relative positions of the routes of course passed them all, (except No. 1, which was abandoned because of its excessive length of tunnel,*) and enables us to determine very nearly the lenth of feeder for any route. Deferring at this time the more detailed account of the difficulties opposed to a feeder in its progress from the Greenbrier River to the summit-level of the second route — which as yet is conspicuous for its advantages over the other routes — it may *To recur to the reasons why but one experiment was made from Dunlap’s to Sec- ond Creek. The great length of tunnel by the first route could not be diminished by any route to Second Creek, for the mouth of Cove Creek (the next tributary to Dunlap’s Creek above Fork Run) is 600 feet above the base-mark, and the rise in Cove Creek being greater than in Dunlap’s Creek we should, by a route through Cove Creek, have to commence tunneling at even a greater distance from Second Creek than by the first route. Indeed, a route from Dunlap’s to Second Creek would be impracticable at any rate, in consequence of its distance from Greenbrier River and of the character of the intermediate country, (p. 764.) 768 REPORT OF THE CHIEF OF ENGINEERS. sujffice our preseut purpose to state tliat the spur \thicb was spoken of (page 766) as projecting from the Alleghany to the mouth of Howard’s Creek, or which lies between Howard’s Creek and the Greenbrier River, constitutes a great objection to every route but the fourth. So circuitous a route for the feeder as that around the extremity of the spur might, and probably would, involve such losses in its course as to be fatal to the practica- bility of the project, while the alternative of pa^^siug the spur in the direction indi- cated by the near approach of two opposite tributaries to the north fork of Howard’s Creek on the one side, and to Anthony’s Creek on the other, however it may diminish the chance of an inadequate supply of water, or may assure us of a probable super- abundance, is fraught with difficulty. The greatest depression in the ridge intermediate to Howard’s and Anthony’s Creeks is at the sources of the two Tributaries just alluded to, through which it had been hoped the feeder might reach the valley of Anthony’s Creek by an open cut, or a tun- nel of moderate length. But it was found to be 1,096 feet above the base-mark, and the slope on either side is so gradual, that limiting the cuts to 35 feet depth, a tunnel of 5 miles and 1,620 yards will be required, (p. 772.) Were the summit-level 50 feet or even 100 feet higher than has been assumed, still the length of tunnel, to pass this ridge, would be, in the one case, 4 miles and 290 yards, and in the other, 3 miles and 533 yards. But the considerations which hereto- fore induced us to limit the elevation of the summit level to 700 feet caonot safely be waived; and therefore, in selecting the location for a feeder, the disadvantages of a tunnel of 4 miles and 1,620 yards, by one route, are to be weighed with the uncertain- ties attending the very greatly-incr' ased length by the other. In comparing the different routes for a canal, however, and the practicability of each, we shall coniine our virw' to the shorter route for the feeder, and on that supposition the following table comprises their principal characteristics : TABLE. Xo. of route. £ c o 6 Cut west of the Allo- ghaiiy. Length of tunnel. Length of the summit- level. Distance from base- mark to summit- level. Distance thence to niouth of Howard’s Creek. Total distance from base -mark to mouth of Howard’s Creek. t>5 n ’Hi laced in favorable circumstances for retain- ing water.” APPENDIX V. 771 may omit to mention, snch as the lenj^th of each level, the lift of each lock, the trans- verse as well as longitudinal slope of the ground, including the works, tlie nature of the soil, rocks, &c., the height of the freshets. It will remain to me, however, to pre- sent a connected and more general view of the whole route. To facilitate a perspicuous arrangement of the subject, let the first section extend from the James to the Greenbrier River; the second section include that portion of the canal along the Greenbrier; and the third section that portion between the mouth of Greenbrier and its teruiina ion at the foot of the Great Falls of Kanawha. First section . — The minuter surveys incident to the locations included in this section, while they test the accuracy of the experimental surveys, develop results somewhat different. The elevation of the summit-level, for instance, was established at (594 feet instead of 700 feet as heretofore assumed, and the depth of cutting at either end of the tunnel through the Alleghany was extended to 50 feet, instead of 35 feet, to which, in comparing the lengths of tunnels by the different routes, we had limited it. The rea- sons for changing the elevation of the summit- level are not, perhaps, important, but the depth of cutting was extended to 50 feet because it was found that the tunnel would be materially shortened ; illustrative of which fact I would add, that on the western side alone, the length of the tunnel is diminished 533 yards by extending the cut to the depth of 50 feet. Dividing this section into three parts, the first subdivision will comprise the summit- level and all the works belonging to it. The length of the summit-level is 4 miles 789 yards, which distance includes a short basin of 98 yards in the valley of Fork Run ; a deep cut at the eastern end of the tunnel of 458 yards; a tunnel through the Alle- ghany of 2 miles and 1,120 yards; a cut at the western end of the tunnel of 1 mile 177 yards, and a basin from the termination of the latter cut, which extends 69S yards. The profile of the ridge immediately over the tunnel indicates frequent great de- pressions, where it crosses the ravines of the head branches of Fork Rnu and Tuckahoe, although between those depressions the ridge sometimes attains a very great height above the summit-level. But the necessity, and to what extent, of sinking shafts in the construction of the tunnel, where those depressions do not exist, will be better seen by designating their position in the following tabular form: 1 2 3 4 5 6 7 8 9 10 11 Num’oer of depressions. Distance fi om the euci of the deep cut at the east- ern end of the tunnel. Height of depres- sion above the summit-level. 1 Greatest height of inte I- mediate ground. Distance from each depression to the oue preceding. Mites. Yards. Feet. Feet. Yards. 0 50U 160 245 500 0 740 166 203 240 0 953 126 249 213 0 1,183 180 230 230 0 1, 516 300 374 333 1 353 276 578 597 1 773 4.58 5'0 420 1 1,017 515 605 244 1 1,677 205 706 660 2 500 130 36S 583 2 1, 120 50 384 620 End of tunnel. In relation to the probable formation of the ridge through which the tunnel would pass, of course we cannot speak with any certainty ; but appearances at and near the surface in the vicinity would incline us to expect the interposition of com[)act sand- stone the greater part of tlie distance. The cutting at either end of the tunnel, it is thought, will prove as favorable for its depth as could reasonably be expected any- where; it will probably consist principally of an argillaceous slate and sandstone of slaty structure, intermixed with a soil (nearer the surface) of sand and clay. Feeder from the Greenbrier River to the summit- level . — The location of this woik does not vary materially from that pursued by the experimental surveys. Its length has been diminished to 31 miles 130 yards, but it still involves a tunnel of 5 miles and 200 yards. The same formation will doubtless be encountered by this tunnel as by that through the Alleghany Mountain ; and the cutting, at either end of tho tunnel, will he similar to what is expected in the deep cuts of the summit-level. The height of the ridge above the surface of the water in the feeder, at different points from the beginning to the end of tae tunnel, is shown in the following table : 772 REPORT OF THE CHIEF OF ENGINEERS. Number of depressions. Distance from the end of the deep cut at the beginning of the tunnel. Height of depression above the surface of t he water in the tunnel. ^ 3 C 5 9 'Z o £ Z u Distance from each depression to the one preceding. Miles. Yards. Feet. Feet. Yards. 1 1, 260 100 262 1, 260 2 1 690 153 1!;9 190 3 1 1,006 204 253 316 4 1 1,374 211 351 368 5 2 409 298 547 795 fi 2 799 373 424 390 7 2 1,066 419 550 267 8 2 1, 426 356 604 360 9 2 1, 622 482 529 196 10 3 52 513 599 190 11 3 362 772 804 310 12 3 675 718 799 313 3 888 733 747 213 14 4 15 703 929 8s7 15 4 155 712 777 140 16 4 435 852 898 280 17 ... 4 735 897 954 300 18 4 1, 235 585 ( Descend- ( 500 19 5 200 35) ing. \ 725 The tiiiiuel begins and ends where the cutting is about 35 feet deep. It will have been observed that the height of the ridge above the tunnel far exceeds the height of the depression spoken of, (page 768,) and the observation leads to the inquiry it it might not be better to give the tunnel the direction of those valleys which head in that de|)ression, instead of a perfectly straight direction through the ridge. In that event the greatest height of the ridge above the tunnel need not exceed 300 feet, and the prolile will show that on both sides the ground rapidly falls to within 150 feet above the tunnel. We could conform to the direction of the opposite valleys, and continue the tunnel straight, or very nearly so, till it reached the valley at Anthony’s Creek, wdien one turn would be necessary. Its length, under those circumstances, would be 5 miles and 950 yards, or it would be but 750 yards longer than if it were perfectly straight. I am, however, rather inclined to think it fortunate that such appalling difficulties as would obstruct the progress of a straight tunnel can so easily be diminished than to hesitate in recommending the longer tunnel because of its increased length ; and I would there- fore add, for the w'hole length of the feeder, the difference between the lengths of the two tunnels, to 31 miles 200 yards. Lieutenant Cook’s report will furnish every other fact connected with the feeder from Greenbrier River, or relative to the reservoir on thar. river. The feeder from Anthony’s Creek he has omitted to mention intersects the main feeder in the 11th mile from the summit-level; it is one mile and 1,509 yards long, and passes down the right shore of Anthony’s Creek without any difficulty. The dimensions of the dam at the head of this feeder are shown on the map. SECOND SUBDIVISION. This extends from the eastern end of the summit-level to the James River, opposite the town of Covington, and includes 19 miles and 73 yards of canal, and 692 feet of lock- age, to the surface of a basin suitable to the reception of the canal-boats. For the first 2 miles and 1,072 yards from the summit-level, or that included in the valley of Fork Run, we unavoidably fall 264 feet; preserving, however, with some ex- ceptions, in which resort is had to contiguous locks, a succession of short levels, con- nected by locks of 8 feet lift. The descent from the first to the second level, and from the seventh to the eighth level, was effected by two contiguous locks; while the third and fourth levels are united by five, and the tenth and eleventh levels by four, contiguous locks. The actual descent of the valley in positions where it was essential to conform as much as possible to that descent to avoid rocky and precipitous slopes, which w^ould else have been encountered, or, in other cases, expensive and deep embankments over broad and deep ravines — in fine, a regard to the various circumstances of the ground — sometimes recommended a combination of locks in preference to a succession of levels. Even with such a disposition as was made of the locks, the canal down the valley of Fork Run passes along a very steep slope, varying from 12° to 38°, till within about f of a mila of Dunlap’s Creek, when it occupies a flat. APPENDIX V. 773 For the remainder of this subdivision (16 miles and 761 yards) or that in the valley of Dunlap’s Creek, the canal was located under much more favorable circumstances, the fallbeinjy much less in a f^iven distance and the valley much wider. As I have re- marked in another part of this report, bottom-lauds alwsys present themselv(*8 on one or the other side ; and although these alternate with bluffs and occasional cliffs, which have more than once induced us to cross the stream, neither the lengths nor heights of the aqueducts would be such as to render them very expensive. It may be satisfac- tory, however, to observe that, whatever difficulty may have presented itself on one side, the creek hasuever been crossed till, from an actual experiment on both sides, it seemed proper to do so ; and the field-uotes may therefore be referred to for the precise motives which determined the location adopted for the canal. From the mouth of Fork Run to its termination opposite Covington the canal is divided into 50 levels, connected with each other, as heretofore, by locks of 8 feet lift, except in three instances, where the descent is made by two contiguous locks. The descent from the last level to the basin proposed at the end of the canal is effected by a lock of 12 feet lift. The excavation along Dunlap’s Creek will, in general, be of the easiest kind, through bottoms or along hill-sides of sand and clay, sometimes mixed with argillaceous slate. The bottoms aie uniformly comx^osed of a vegetable mold about a foot thick, resting on a bed of clay. Linitstone is found in the valley of Dunlap’s Creek, opposite the canal, near Crow’s tavern, and in several other places ; but it was not observed to in- tersect our line ; in fact, we have little apprehension that the canal would experience more than the ordinary losses incident to Similar works favorably situated for retain- ing water. Provision was made for the introduction into the canal of the supply afforded by Dunlap’s Creek ,* it is to be very easily effected by a short feeder of but 610 yards’ length, from a dam barely high enough to divert the course of the stream. The situa- tion of the dam is not far below the mouth of Fork Run, and the level into which the supply is introduced is 2 miles and 1,718 yards from the summit-level. THIRD SUBDIVISION. This comprises a succession of short levels, connected by single locks of the uniform lift of 8 feet, and extends from the western end of the summit-level through the valley of Howard’s Creek to the lefo bank of Greenbrier River. It includes a ilfstance of 8 miles and 155 yards and a descent of 216 feet. It is unnecessary, however, to dwell upon this subdivision ; it presents no difficulty deserving of particular comment, and all its features are sufficiently developed on the map. The supply from Howard’s Creek, if at all needed, may, perhaps, with most facility, be brought into the sixteenth level, which begins just below a very rapid part of the creek. A feeder of but 35.5 yards’ length, with a dam 20 yards long and 7 feet high, would be sufficient. A review of the first section of the canal affords the following summary of some of its principal features Distance. No. of locks. Ascent. Descent. .Tanifis Tlivftr to summit, -Iftvftl Miles. 19 Yards. 73 86 Feet. 692 Feet. Summit-level 4 789 Summit-level to Greenbrier Eiver 8 155 27 216 Totfl,l 31 1, 017 113 692 216 Feeders. Length. From Greenbrier Fiver to summit-level From Anthony’s Cieek to main feeder Miles. 31 1 Yards. 130 1, 509 610 From Dunlap’s Creek to twenty -seventh level of canal From Howard’s Creek to sixteenth level of canal 333 Total length of feeders 33 882 I have omitted to advert to the diminished contents of the reservoir on Greenbrier River resulting from Lieutenant Cook’s selection of the site fora dam, (see his rejmrt, page 787,) because no doubts as to the adequacy of the supply have arisen from that circumstance. For it will be perceived that, by increasing the height of the dam 774 EEPORT OF THE CHIEF OF ENGINEERS. to 60 feet, we can reserve 1-1,000,000 cubic yards, instead of 13,060,584, on which latter quantity our former leasoniiijr was predicated, and which, it has been showu, would leave a superabundance beyond all the wants of the canal. Second iecfion . — The first section terminating but 33 feet above low-water mark, viz, the surface of canal, it is thought advisable to ])rolong that level across the river; for the height to which freshets sometimes rise might else endanger the security of the aqueduct. The length of the aqueduct under those circumstances would be 167 yards, at the end of which wedesceud by two locks, of 8^^ feet lift each, to what Lieutenant Hazzard terms his first level. An additional supply of water may with great ease be intro- duced into the canal on either side of the river, even at the elevation of the aqueduct. A feeder on the left shore of but 1,260 yards in length, from a dam 13 feet high and 108 yards long, would effect the object. But I think it to be preferred that the new supply be admitted just beyond the aqueduct, from which to a dam of even less elevation a shorter feeder along the right shore would answer equally well. Arrived at the first level beyond the a(iueduct, the location of this section was con- tinued, under various circumstances, to its termination at the mouth of Greenbrier River. The cliffs which impinge upon the stream, and which at times the canal un- avoidably encounters, may rather be considered exceptions to the otherwise generally favorable nature of the ground, than as characterizing this section as at all remark- able for the extent of the obstacles to its eas3’’ execution. Its length is 49 miles and 151 j^ards, which distance is subdivided into 36 levels, (the aqueduct being included,) united by locks of the uniform lift of 8 feet, excepting the two locks of 84 feet lift at the end of the aqueduct across the Greenbrier. The canal therefore descends 297 feet in its progress to New River.* The only tributary streams of any consequence crossed by this section of canal are Mill Creek, Muddy Creek, aud Hunger’s Creek; the first requiring an aqueduct 100 feet long aud 17^ feet high, and the two others aqueducts of but 50 feet length and 17| feet Ijeight. It does not appear advisable to introduce either of them into the canal; for every facility exists for obtaining an abundant supply from the river, to which we must necessarily resort. Suitable sites for dams acioss Greenbrier very fre- quentl^' occur, but their position aud dimensions will readily be seen on reference to the maps. On the left, or east side, as we descend below Howard’s Creek, the number and mag- nitude of the tribut iiy streams are also so limited, that we are led to ascribe the in- creased size of Greenbrier River, observable in its progress'to New River, rather to the existence of numerous springs which rise within or near its bed than to contributions from more distant sources; Second Creek, perhaps, being the only stream from the left which materially adds to the volume of the river. Third section . — This extends 67 miles and 779 yards, and includes a total descent of 762 feet to the surface of a natural basin below the Great Falls of Kanawha River, and it is the peculiar character of a portion of the valley included in this section which would present the most appalling difficulties to the construction of a canal. For 45 miles, or to Bowyer’s Feriy, obstacles, such as are almost continuous between Bowj^er’s Ferry and the Gaule.y River, are confined to but a comparatively small por- tion of the distance, but below Bowj’^er’s Ferry it would be only at great expense that the canal could be protected against the impetuosity" of a current almost resistless dur- ing the swollen stages of the river. This is exemplified in the fact that freshets sometimes attain the great height of 35 aud even 50 feet in some places, within whose utmost reach there could be no security to the durability^ of any work, and from the nature of the intermediate space between low-water mark aud the cliffs, which rise nearly perpendicularly for several hundred feet ; for notwithstanding, even below Bowyer’s Ferry, there is generally room enough for the canal between low-water mark and the cliffs, and that the interval may never be wholly overflowed, it is occupied by mssses of huge rocks exhibiting a surface almost too irregular to be defined. These rocks, however, may be converted into a kind of bench, on which the canal could be sustained beyond the reach of freshets, and eventually might facilitate ifs construction ; but the necessity of high and extensive walling would still exist, and the total absence of any suitable material for puddling the canal would be irremediable except at the cost and trouble of procuring it from a distance, unless, indeed, clay of a proper consistency be obtained near the verge of the cliffs, when this latter source of expense may" be very much diminished from the ease with which it might be deposited at convenient intervals along the line. The rocks at the foot of the cliffs have every appearance of having at some period been precipitated from different points of the cliffs, and, although ages may have since elapsed, it must be obvious that it would be impossible to secure any work at the base of the cliffs against the destiuctive effects of a similar occurrence. It is not con- *The Great Kanawha River is more familiarly known as New River above its conflu- ence with the Gauley River. APPENDIX V. 775 ceiv'ed, however, that much importance is due to this suggestion, since there is little proLability that such masses as those alluded to would in future be detached, unless by some extraordinary convulsion of nature beyond what we have reason to anticipate. Above Bowyer’s Ferry the valley is much wider, and, although the canal would fre- quently occupy steep slopes, such difficulties as exist below are not to be apprehended either from the heights of freshets or the perpendicularity of the banks. No trace was perceived of the water ever having risen more than thirteen feet above its ordinary channel until within a few miles of Bowyer’s Ferry, and, with the exception of a few places designated by Lieutenant Hazzard as requiring walling to protect the canal, there is comparatively but little difficulty in sustaining the line beyond the reach of freshets. The tributaries to New River, between Greenbrier and the Gauley Rivers, are few and uuimportaut; and the Great Falls of Kanawha, two miles below the Gauley, des- ignates as well the commencement of an entirely different country from that above as the termination of our ideal section. Tug Falls, Buffalo Falls, Richmond Falls, and the Great Falls, with an almost con- tinuous rapid, when perpendicular falls do not occur — in all, constituting a spectacle the sublimity of which can scarcely be surpassed — characterize New River below its confluence with the Greenbrier as an impetuous and almost resistless torrent. And the general aspect of both sides of the river, with the wild features of its valley, can offer few temptations to the intrusion of civilized man. B it the Great Kanawha ma- jestically pursues its placid course through a rich and fertile valley, and is said to present but few obstructions to a perfect navigation from the falls to the Ohio River. The general structure of the country below Greenbiier River is based on sandstone (gray and red) or a compact limestone, and coal of an excellent quality exists, in ex- haustless quantities, from Sewell Mountain westward. Having now sufficiently reviewed the several sections in detail, the following sum- mary of the whole route will conclude the subject which has heretofore occupied us : SUMMARY. Distance. Lockage. James River to the Greenbrier River Miles. 31 Yards. 1, 017 In feet. 908 Thence to New River 49 151 297 Tlienee to the ha, sin below the Great, Ea.lla C7 779 762 Total 148 187 1,967 ROANOKE AND KANAWHA CANAL. On completing the surveys relating to a canal from James River to the Great Ka- nawha, those having reference to other objects enumerated in my instructions were immediately undertaken; and although a detailed report on the Roanoke and Ka- nawha Canal is unavoidably postponed till the drawings illustrative of the su rvey shall be more advanced, a brief summary of the more important facts may, at this time, be satisfactory, as it will suffice to show the great facility with which those riveis may be united. The experimental surveys, which extended along the Alleghany Mountain, from the southern source of one of "the branches of the North Fork of Roanoke, beyond Christ- iansburg on the north, to the South Fork of Elliot’s Creek on the south, indicate the existence of frequent great depressions in the dividing ridge, or rather a general de- pression of the Alleghany Mountain, in its course between the waters of the Roanoke and those which, flowing westward, empty into Little River. But the point which imposingly presents itself as offering superior advantages for the passage of the Allegheny by a canal is at the sources of the North Fork of Elliot’s Creek, a tributary to the South Fork of Roanoke, and “ Green Head Branch ” of Meadow Creek, a tributary to Little River. An abundant supply of water might be obtained from Little River, even were our summit-level of the same elevation as that of the very top of the mountain, in the depression alluded to ; but apprised of that fact from the results of our experimental surveys, sufficient considerations recommended, notwithstanding, a location of the canal in the following manner: The summit-level commences on the east side with a cut, which in 1,320 yards attains its greatest depth of 30 feet, (above the surface of the canal,) and terminating on the west side 144 yards beyond the end of the cut, includes, iu all, 1,404 yards. The canal then descends, through the valley of Meadow Creek, to Little River, and 776 REPORT OF THE CHIEF OF ENGINEERS. thence along Little River to Nevr River, under circumstances as favorable as might be expected, no difficulties occurring worthy of comment at this time; the distance from the summit-level to the latter point being 10 miles 769 yards, and the fall 288 feet.* From the mouth of Little River it is prolonged on the right bank of New River to its intersection with the James and Kanawha Canal ; and to the mouth of Greenbrier, for most of the distance, the canal occupies very iavorable ground through extensive llats which border on the river. These flats, however, are not continuous, and frequently, for short di.stances, we must encounter cliffs which impinge upon the streams ; but, with the remark that the dif- ference is rather in favor of the valley of New River, we shall not create an erroneous impression if we refer to the location down the Greenbrier as the standard by which we may judge of that between Greenbrier and the mouth of Little River. Returning to the summit-level, the descent on the eastern side, until we arrive within half a mile of the main valley of the North Fork of Elliot’s Creek, is sucli as to neces- sitate a succession of short levels, and the canal falls, by locks of 8 feet lift, 120 feet in but 1 mile 485 yards; the next 11 miles 115 yards bring us to the mouth of Elliot’s Creek, and comprise 62 le' els and a descent of 496 feet ; whence, through the wide and fertile valley of the South Fork of the Roanoke, the location extended to the forks, where the fleld-operations of the brigade terminated for the season. The distance from the mouth of Elliot’s Creek to the termination of the canal, near • the forks of the Roanoke, is 9 miles 1,320 yards, and the fall 200 feet.t The canal, there- fore, from the mouth of tbe Greenbr ier River to the latter point, includes an ascent of 657,7 feet, in a distance of 94 miles 106 yards, (the distance from the Greenbrier to the mouth of Little River being 83 miles and 1,097 yards, and the rise 369.7 feet,) a sum- mit-level of 1,464 yards, and a descent in 22 miles and 100 yards of 816 feet; or the total distance is 116 miles and 1,730 yards, and the lockage 1,473.7 feet. SUPPLY OF WATER. It is the great facility with which this is obtained that so distinctly characterizes the connection in view as one so very feasible in its execution. Little River, on which the summit-level is dependent, was found to yield at its lowest stage nearly 100 cubic feet per second, when its supply can be commanded by a feeder but 9 miles and 1,225 yards long. IMlot Mountain, howevfr, which lies between Little River and the summit-level, preserves such an elevated character, that the feeder can only pass it by a tunnel of 1 mile and 290 yards’ length. We might, indeed, wind along the slope of the mountain, but it would so very greatly increase the length of the feeder, that there can be no hesitation in preferring a tunnel. In other respects, the construction of the feeder would be attended with little difficulty. The supply from Little River alone would be ample to all the wants of the canal from New River to the forks of the Roanoke ; but a reference to the report of Lieuten- ant Fessenden, which exhibits tbe discharge of the two branches of the Roanoke, of Elliott’s Creek, of Meadow Creek, and the increased size of Little River, in its progress to its mouth, will show the abundance of its other resources. Of the practicability, therefore, and comparative ease with which the Roanoke may be united to the Kanawha by means of a canal, the brief statement of facts which we have given will have been sufficient to dispel all doubt, and, on that conviction, we might forego all further discussion of the subject, but that a few other remarks natu- rally suggest thfmselves as neither uninteresting nor entirely irrelative. If the connection be regarded merely as an avenue for the trade of the Ohio Valley, its importance in that light may'doubtless, at some future day, elicit tbe effort to overcome those obstacles between Bowyer’s Ferry and the Great Falls, which consti- tute almost the sole impediment to its comparatively easy execution. But should the magnitude of these obstacles be considered sufficient forever to discourage the enter- prise of a nation, the importance of the connection, although diminished, is yet con- spicuous. New River, for perhaps 100 miles above the mouth of Little River, it is said, traverses a country rich in mineral and agricultural products, and its navigation may prove easily susceptible of great improvement, while the direction of Reed Creek, a tributary of New River, above its confluence with Little River, is pariicularly spoken of as promising facilities for effectuating a connection with the Middle Fork of the Holstein. Of this I cannot speak from personal observation, nor from information entirely au- thentic, yet the concurrent opinions of individuals acquainted with the country would * The bench at the mouth of Little River, on the right bank, is 290 feet below the summit-level, and 16 feet above low- water mark. tThe 1 euch-mark near the mill at the forks of the Roanoke is 831 feet below the summit-level. APPENDIX V. 777 seem to warrant a belief in the practicability of a canal from New River to the Hol- stein ; and the relative situation of the Tennessee and Alabama Rivers, as delineated on every map, renders it by no means improbable that they, also, mif^ht be united. Thus it is possible, and I mi^ht even say within the scope of probability, that a canal from the Roanoke to the Kanawha may at some future day be re<>jarded but as the last link in a chain of inland communication from the Gulf of Mexico to the Chesa{)eake. From “a due investit;atiou of the hydroosses8 similar advantages for uniting either the James or Roanoke River with the Kanawha by means of a canal. The country between Knapp’s Cn ek, a branch of the Greenbrier, and Back Creek,^a tributary to .Jackson’s River, and between Craig’s Creek and Sinking Creek, opposite tributaries to Jackson’s and New Rivers, was said to atford some facilities for a canal from the James to the Kanawha River, and was, in consequence, reconnoitered. But it )s apparent that no route in either of those directions can present any claim to further examination, (see Lieutenant Dillahunty’s report, page 800,) and the conclusion is equally obvious, from our pref^ent knowledge of the country, that the James River cannot be united by a canal with the waters of the Great Kanawha by any route below the valley of Dunlup’s Creek, unless, indeed, as is highly probable, a canal from James River through the valley of Catawba Creek be found practicable to the R lauoke. The examination, however, of the country iuTermediate to the Roanoke and .James Rivers, with reference to a canal or railroad, is one of those objects enumerated in my instructions whicb as yet, trom the want of time, have been omitted. Of the practicability of a railroad from the Janies or Roanoke River to the Great Falls of Kanawha there cannot reniiin a doubt; and the surveys which have been made of the intermediate country will, in general, furnish ample means for deciding on the most proper routes. A single glance at the topography precludes all hesitation in selecting, for a railroad from .James River, some one of the routes surveyed for a canal in preference to a more direct route over the high and numerous ridges which intervene between the Greenbrier and Gauley Rivers. To be more explicit, however, with such deviations only as would result from the different characters of the two works, the route w hich has been adopted for a canal as far as the mouth of Muddy Creek wmuld be pursued for a railroad. A doubt is sug- gested as to the best direction for continuing the route beyond that point only because it is possible that a route through the valley of Muddy Creek and across to Meadow River, a tributary to the Gauley (from the circumstance of there being but one ridge betw'eeu Muddy Creek and Meadow River) may be found to possess advantages which may bring it in competition with a route through the valleys of Greenbrier and Kanawha Rivers. But supposing it to pursue the same route as the canal, as is thought most probable, theleugtnof a railroad from Covington to the Great Falls of I^auawha would lie about 148 miles, with a rise and fall of 2,763 feet, on the supposition of passing the Alleghany without a tunnel. The surveys from the Roanoke do not altogether determine the best route for a rail- road to New River, beyond which it w’ould, of course, continue down the valley. But the discovery that the Alleghany is nearly as low at one of the sources of the North Fork of the Roanoke near Christiansburg as on the route proposed for the canal, with the fact of the direction of the Norih Fork being such as to afford a shorter route than cue up the valley of the South Fork, makes it more than probable that a railroad from the Roanoke would cross the Alleghany in the vicinity of Christiansburg. However, whatever may be the deviation from jOur location of the canals by the sub- stitution of railroads in their stead, they cannot be so material as to interfere with those general considerations which may determine, from what has now’ been stated, the comparaiive mt^rits of railroads and canals as the means of uniting the James or Roanoke River with the Great I^^auawha. Having now" briefly adverted to the operations of my brigade subsequent to the com- pletion of the surveys relating to a canal from the James to the ICanawdia River, in concluding this report I may be permitted to advert to the causes which have delayed its completion beyond the period at which the Department may" have had reason to expect it. A reitort simply on the James and Kanawha Canal could have l)een pre- sented at the commencement of the present session of Congress but for circumstances entirely unfore eeu. I allude to the experimental surveys in relation to the Baltimore and Ohio Railroad, which were undertaken, with the assistance of my brigade, as late as the latter part of November and continued through a considerable portion of the inclemency of w"inter, when heretofore it has, in all cases, occurred that the maps and profiles illustrative of the operations of the smnmer-season have immediately succeeded our return to winter-quarters ; and since the completion of those exj)erimental surveys I have, from an impression that it would be more satisfactory to present an equally detailed report on both the James and ICanawflia and Roanoke ami Kauaw'ha Canals alternately, employed myself, till within a few days only, on the former subject and on 778 REPORT OF THE CHIEF OF ENGINEERS. a report on the Baltimore and Ohio Railroad, awaiting a more advanced state of the drawings to collate the facts essential to an equally detailed report on the Roanoke and Kanawha Canal. Time has not sufficed, however, to execute fully my intentions. Which is most respectfully submitted by, sir, your very obedient servant, Wm. G. McNkill, Captain United States Topographical Engineers. Topographical Ofi^ice, Gcorgetoicn, March 24, 1828. APPENDIX. CONTAINING REPORTS FROM LIEUTENANTS COOK AND KAZZARD, ILLUSTRATIVE OF THE LOCATION OF THE JAMES AND KANAWHA CANAL, WITH REPORTS FROM LIEUTENANTS DILLAHUNTY AND FESSENDEN ON THE SUl'PLY OF WATER. Georgetown, D. C., March 10, 1828. Sir: I herewith submit to you the results of those operations which, in compliance with your instructions of May 3, 1827, occupied the party under my command until those instructions were fulfilled. They relate to the location of a line of canal from Covington, on Jackson’s R»ver, to the Greenbrier River, by the route designated in your instructions to me, as well as the location of every work connected with it, such as feeders, reservoirs, dams, &c. Dividing the line of canal into three parts, of which the first includes the summit- level, the second the portion east of it, and the third the descent to the Greenbrier River, I proceed, to enumerate the details of each subdivision. FIRST SUBDIVISION. This includes a tunnel of 2 miles and 1,120 yards, two cuts of 50 feet each in depth and two basins; together constituting the summit-level, which comprises 4 miles and 789 yards. Its elevation is 694 feet above the base-mark. Further details in relation to this subdivision are omitted, because they have already been furnished you. second SUBDIVISION. To avoid the constant repetition of facts which so frequently recur, a table has been formed to present at one view the details belonging to each level of this subdivision. To render it entirely intelligible, by the first level is meant that to which we descend ftom the summit-level by the first lock; the number of locks uniting adjacent levels refer to the number between each level and the one preceding ; the length of the level is made to include the space occupied by the locks uniting it with the adjacent level. The length and height of acqueducts and culverts are in feet, length of walling in yards, aud the height in feet above the bottom of the stream which it rests in. Defails heJonging to each level of the second stthdivision of the contemplated canal from Covington, on Jackson's Biver, to the Greenbrier Biver. APPENDIX V, 779 3 S S fall Sooocooocoooooooccooooooooooooo o o C'O'B'S'C’w'C'C'C'O • • •^'C-C'T3 tjTJ< : : : ^ ^ I 4 ;0 O O • o CO 'CD 43 : : : : : : •e 5 tDO[JO .laqran^ HOrHi-HrHCNr-l.-l'^TH— ' — rHrHrHCJi I 0< rH r-l »-( o u r |liSiS2iiSli2i2s2riiSs3lgls22i?° I i S iSilS? S o O O O O O O O O O O O O O O O OO OO O O O O O O O O O O ri o o o ooocoo J • 18 A 9 [ JO aaqmnx ^7ir^^^-^r^ODOO-3.«;^ooj^;«^.o-cjg=^«or;jaooo- ^ ^ ^ ^ ^ ^ o Details belonging to each level of the second subdivision of the contemplated canal from Covington, ifc. — Coutiniiecl. 780 EEPOKT OF THE CHIEF OF ENGINEERS. *S>[ 00 | jaqcun^ O O 'O o o H o O O CO o o o o o o o o O CO O O O 05 O O O CO'^CO Qi to r-irHt'- CO lO’^COQCr- — T-»rMr-l r- CO CO o CO ^ o CO 00 CO Oi CO CM Oi o o o o ooooooooo ooooooooooo C 50 t-» W'^iccor-QOCiO— cc^iocor-xcio— ‘C^co -Vioo o iOOOiOlOlOOCOCO CO COCOCOCOCOCOC0 4-i-t-t>* • 19 A 0 I JO aequinx rr TT rr •n' 0 M 2 feet. APPENDIX V. 781 From the foregoing table it may seem that some objectionable features attend the location of the canal, such as the occurrence of contiguous locks in live places, the crossing of Fork Run and Dunlap’s Creek by aqueducts, and the deep cuttings in the sixty-second level. The reasons which influence me in each instance will therefore be detailed. The levels along Fork Run, from the descent of its valley, necessarily had to be very short. The length of the first level was greater than could have been the average by single locks, and, to avoid contiguous locks, on that account alone we should have terminated it. But a single lock Avould have brought the next level in contact with unfavorable ground, necessitating eitlitr a circuitous route around a ravine or an embankment across it; and, to diminish the difficulty, we descended at once by two locks. At the end of the third level the same reasons for five contiguous locks obtained, but to a greater degree; the ground was exceedingly steep at the elevation at which we were above the stream, and a wider and deeper ravine was before us. The seventh level w'as teiminated by two contiguous locks, that the eighth level might occup3^ the most favorable ground, and the tenth level was terminated by four contiguous locks to avoid a high and precipitous cliff of sandstone. Thus far the left shore was the preferable one, but thence to the mouth of Fork Run the right was decidedly better, because of the frequent occurrence on the left shore of steep and rocky ground. We therefore, on arriving at the end of the tenth level, descended, in order to cross as soon as possible; and to do so with the shortest and lowest aqueduct, we resorted to two locks. In the valley of Dunlap’s Creek the descent from an upper to a lower level is only in three places effected by two contigu- ous locks. In the first two cases it was to shorten the distance to avoid very steep ground and to take advantage of very favorable flats. In the last case it was to dimin- ish the height of walling necessary to pass the cliff, which occurs in the seventy-first level. The reasons for crossing Fork Run in the nineteenth level have been already given in the second instance near the mouth. It was crossed to save the distance to a favor- able point for crossing Dunlap’s Creek above the mouth of Fork Run, to say nothing of the ground on the left shore of Dunlap’s Creek being decidedly jireferable to that on the right for nearly 2 miles. Dunlap’s Creek was then crossed, because the expense of crossing Brush Creek, Lick Creek, and sundry smaller runs which enter on the left, would, in itself, have been more than the expense of crossing and even recrossing Dunlap’s Creek; and, besides, the ground on the right for more than 5 miles was known to be much better than that on the left. The canal subsequently crosses Dunlap’s Creek four times in its progress to the mouth of the creek. The high and extensive cliffs which occur on the right shore between the fifth and seventh miles, it was thought, entitled the left shore to the preference during that distance; but the entrance of Ogley’s Creek just below the fifth mile, and the continuance of very steep and rocky ground on the left, again renders it expedient to gain the right shore. The last two crossings of the creek are incurred rather than wind around the bend of the stream, on the right of which there is a deep ravine, and very steep and rocky ground: and as, agreeably to your instructions, the canal was, if practicable, to follow nearly the direction of the turnpike beyond that bend, one crossing, to avoid the rocky ground alluded to, necessitated another to conform to your instructions, since it was found practicable and altogether more expedient to follow the direction of the turn- pike than the very circuitous course of the creek. In evidence of this fact the alterna- tive, which is prefer/ ed, of crossing in the depression through which the turnf>ike is located, instead of following the valley of Dunlap’s Creek, saves one thousand and thirty-three yards of canal along very unfavorable ground; at the expense, however, of the deep cut in the sixty-second level. As to the termination of this subdivision, you remark in your instructions to me, “ Since it may reasonably be expected that, if ever a canal be made across the Alle- ghany Mountains to James River by the route to be surveyed, it would be continued down the river, or the navigation of the river would be so far improved as to admit the passage of boats such as would be adap ed to the canal, you will, to i/rovide for either contingency, so conclude your survey on reaching James River that the length and height of an aqueduct to reach Covington may be determined, (as in the event of a continuation of the canal it is through that town that the chief engineer of Virginia has recommended its location,) besides so locating it that in the event of its termination on this side, and the improvement of the navigation of the river by locks and dams, the lockage, &c., may be drawn to a basin on such a level as would be formed by the construction of a dam of some moderate elevation.” The location exhibited by the profilo of the canal refers to the latter supposition, and the descent from the last level to the surface of a basin formed by a dam 8 feet high, two hundred and seventeen yards below the mouth of Dunlap’s Creek, is supposed to be eftected by a lock of 12 feet lift. The length of the dam is 300 feet. 782 REPORT OF THE CHIEF OF ENGINEERS. On the snppositioD of au aqueduct to cross to the town of Covington, in order to admit the requisite thickness to the arches and be beyond the reach of freshets, (which rise from 5 to 10 feet,) the seventy-first level should be continued across the river; the length of that level would be one mile and eight hundred and thirty-five yards, iucludiug au aqueduct 300 feet long and 21 feet high, counting from the bottom of the stream. The heights of freshets in Dunlap’s Creek vary, of course, with the fall and width of the stream, but they may be said to rise from 5 to 7 feet at most. The general depth of the stream is so trifling that only the low-w’-ater mark has been represented on the profile; the height of aqueducts, walling, &c., is, however, counted from the bottom of the stream. On reaching the valley of Dunlap’s Creek, those coos’derations obtained which you directed in reference to the introduction of a supply of water from Dunlap’s Creek without “the risk of inundation from freshets,” the canal was dropped “so as to recpiire but a short feeder from a suitable site for a dam of moderate elevation.” The section of a dam on an enlarged scale is given on the map ; the length of the feeder is but 610 yards. THIRD SUBDIVISION. From the western end of the summit-level to bench-mark on the Greenbrier River. To this subdivision there is a table annexed, to comprehend precisely such facts as are contained in the table connecteel with the second subdivision. APPENDIX V. g O O O O O O C O O O O O O O O CO o oooococo rg rs r" 'C'Cl'OJ'O'C'w'OJ'CJ -13 ’V > 11 I S ocoo'^. ‘oooooo S i i i :ES sill I i i a I gfi"ffl! «o c: «3 TO o 00 CO 4 •S>100[ JO joi^uiux; = 13 ti > psiSsliiSiisi i iissSsiiSgiiig gooooooc.oooc.oo o o o oo o oooooooo J •| 9 Aa{ JO jonuiuj^ Si wv lO CO t— c »0 CO I' 00 o c — s«?JSi5;-sssi 783 784 REPORT OF THE CHIEF OF ENGINEERS. The foregoing table, with the map and profile, showing the location, renders it unnecessary to dwell long on this subdivision. The reasons for first crossing the creek when it had been determined to cross the Greenbrier River from the right bank of Howard’s Creek were simply because the dis- tance was diminished and more favorable ground obtained on the left side ; and had we not considered these and remained on the right shore our line would have been obliged to cross the Middle and North Forks of Howard’s Creek. In the ninth and tenth levels the canal four times crossed a channel occupied by a part of the stream in time of freshets, but the stream may easily be confined to the main channel by a short dike, and therefore it was not considered an objection to the location. We recrossed the creek in the tenth level for similar reasons ; that is, to shorten the distance, &c., because, as we just now observed, the Greenbrier was to be crossed above the mouth of Howard’s Creek. In the eighteenth level we unavoidably cross the turnpike, and it may be as well to remark that the transverse slope along the nineteenth level, althbugh represented very steep, is terminated by a flat not more than about fifteen feet above the canal; and the cutting represented at the end of the level is the height of a point of bottom-land which we cross because of the precipitous slope of the ground near the creek. The walling is nowhere very considerable, and where it has occurred it was una- voidable. In general, it may be remarked that the canal occupies very fiivorable ground throughout this subdivision. On arriving at the bench-mark designated as the termination of this subdivision, and which is 3 feet below the bottom of the canal at the last level, a feeder was run up the river on the supposition that an additional supply of water might be required. This it was found can be accomplished by a feeder 1,260 yards long through the most favorable cutting, and a dam across the Greenbrier of 108 yards in length and 13 feet high. The feeder intended to introduce a supply from Greenbrier River into the canal was located from the western end of the summit-level, at an inclination of 6 inches per mile, until it intersects the Greenbrier River, or, rather, until a dam 5 feet high was required to turn the water into the feeder. The length of the feeder, by the route sur- veyed, is 31 miles 130 yards, but as it includes a tunnel of 5 miles and 200 yards through the spur of the Alleghany Mountain, which divides Howard’s from Anthony’s Creek, our reasons for not surveying the route, also, around that spur, as you directed, will be given. We shall divide the feeder into such portions that the first and third will include the parts which would be common to both routes; and the second portion, including the tunnel, will then compare directly with that part around the spur, from the end of the first to the beginning of the third part. The first section pursues a level along pretty favorable ground, with the exception of two or three places where deep cuts for short distances must be encountered from the summit-level to within but a short distance of the tunnel, and includes 4 miles and 396 yards. The second section continues on a level but 100 yards, when 1,497 yards are occupied by the two cuts, which are supposed to terminate at 3.6 feet depth at either end of the tunnel; that is, the cut on the southern side is 1,187 yards long, the tunnehS miles 200 yards, and the northern cut 310 yards long; the remainder of this section extends but 740 yards farther to the aqueduct over Anthony’s Creek ; making the length of the second section 6 miles 747 yards. Now, suppose, instead of tunneling through this spur of the mountain, we follow the level, and see what objections there are to this location. From the end of the first section, the level was continued with the intention to pur- sue it until we should arrive, as we did by the route through the spur, at the first favorable point for crossing Anthony’s Creek ; but an experiment of but 10 miles over what we knew was a fair specimen of the remainder induced us to suppose that, had 5 mu been x>resent, you would have considered it, as we did, useless to j^roceed farther. The spur is indented on both sides by innumerable ravines, bounding the small branches tributary to Howard’s Creek or the Greenbrier River; and, as a proof of the great variation from anything like a tolerably direct course necessitated by those ravines, while, by the route of the feeder, following the proper level or inclination, the distance is 10 miles, between the same points, by the road through the valley, the distance is but 5 miles. We know, by former surveys of last year through the valleys of Howard’s Creek, the Greenbrier River, and Anthony’s Creek, from the point where we terminated our exx)eriment to the point at which we should be obliged to ascend Anthony’s Cieek to cross it, the distance could not be hss than 21 miles. But these 21 miles pass over ground no more fiivorable than was that which we had found to double the distance beyond what, by the same standard, without an actual location, we should have consid- APPENDIX V. 785 ered it. We mi^ht, therefore, with little risk of map:nifyin" the length of feeder by this route, add at least one-tliird to the 21 miles, and thus make the saving of dis- tance by the tunnel 32 miles. This excessive length, however, wouhl not alone have deterred us; but, knowing as we did that altjiost throughout these 32 miles the feeder would pass over very steep ground, varying from 25^ to 45°, we concluded the losses of such a feeder, so situated, would certainly leave a supply altogether inadequate to the wants of the canal ; or, in fact, we concluded the other was the only practicable route. It was this conclusion which determined us to improve the means at our dis- posal in the execution of surveys which it was known to bo desirable, if possible, to finish before the close of the season. The third section includes the remainder of the feeder. It begins with the aqueduct across Anthony’s Creek, which is 40 fe-t high and 120 yards long. Different experi- ments were made to ascertain the most favorable place for crossing this creek at a less elevation above its bed. In one case, with an aqueduct of 32 feet high and 170 yards long, it increases the length of the feeder 1 mile; and in another, still higher up, with an aqueduct 20 feet high and 133 yards long, the distance was increased to 3 miles 342 yards. The si‘e at last adoxJted was that at the point lowest down the creek, being the best that could be found. At 487 yards beyond the aqueduct we arrive at the end of the eleventh mile from the summit-level, and all that relates to the remaining part of the feeder will be found contained in the following table: 50 E Table, 786 REPORT OF THE CHIEF OF ENGINEERS. o o o ftp;:; p p O ^ I 5^ CO CO eo 00 to of o' of O^ rH O'! TO irTo" o' o'- CO CO 0» CO 00 co^o''i3 OI -ir. 00 .1 OI LO ■ OOOOOOOOOOOO! o o o o o < ■89{i(n JO aoqinn^ 0» CO •<1< lO C£> t- 00 at at at at APPENDIX V. 787 A reference to the foregoing table shows, as the most unfavorable feature of that part of the feeder between Anthony’s Creek and its termination, the steep slope along which it necessarily is located ; nor did it seem but that this didicnlty would be rather increased tlian diminished b\' keeping a higher level. Tlie two tunnels which occur in the Pith and 17th miles are particularly recommended, in preference to the circuitous course we should pursue in following the level around those points through which we tunnel. For, omitt ing any cutting at either end of those runnels in the one case, while the distance through is but 21 yards, the distance around is 2 miles 1,000 yards, over very rocky and precipitous ground ; and, in the second ease, the tunnel is hut 100 yards, and the distance around 1 mile 738 yards, over very unfavorable ^round. It will be seen that limestone is but very seldom crossed by the teeder ; generally speaking, a com- pact sandstone is found ; rock-excavation will occur in but few places; loose recks, however, are along a great part of the line, beneath which there is a clayey and grav- elly soil. When walling is mentioned, it is suppo-ied necessary on account of the per- pe dicular cliffs ; it, however, may prove better to blast the rock in such places, so as to obtain sufficient width for the feeder without walling. The dimensions of a dam sufficiently high to turn the wa^er into the feeder will be seen better from the map ; its greatest height — for it crosses an island — is 5 feet, and its length is 120 yards, including 30 yards for the width of the island. The position of the dam for the reservoir to be formed was chosen 28,5 yards above the low dam just spoken of, where the valley is very narrow and the sides composed of rock. It might be raised to almost any height, and the relative lengths of dams of 30, 50, CO, and 70 feet high are as follows, (referred to on p. 773) : Dam 30 feet high : length at bottom, 70 yards ; ditto at top, 132 yards. Dam 50 feet high ; length at bottom, 70 yards ; ditto at top, 195 yards. Dam fiO feet higu ; length at bottom, 70 yards ; dit o at top, 212 yards. Dam 70 feet high ; length at bottom, 70 yards; ditto at top, 228 yards. The area and prism of each are given below : Dam .30 feet high ; area, 023,155 square yards ; prism, 2,596,666 cubic yards. Dam 50 feet high ; area, 1,278,400 square yards ; prism, 8,578,148 cubic yards. Dam 00 feet high ; area, 1,737,162 square yards ; prism, 14,000,000 cubic yards. Dam 70 feet high ; area, 2,191,046 square yards ; [trism, 21,000,000 cubic yards. The site of the dam selected the first year, in making the experimental surveys, was higher up the river, (720 feet above the base-m«rk,) but it would overflow a much greater area in proportion to the cubic contents of the reservoir. In every case, however, where there was occasion for the discretionary exercise of judgtnent in the executi(>n of the duties assigned to Lieutenant Fessenden and myself, the alternative to the course adopted by us was never rejected without our concurrent opinion that its comparative advantages could not bring it hereafter in competition. A report of the other objects (the railroad from Covington to the Greenbrier River, and the connection of the Roanoke and Kanawha) which subsequently occupied the party under me is delayed until the drawings shall be more advanced. I am, sir, very respectfully, your obedient servant, WirxiAM Cook, Lieiilenant United States Artillery^ on Topographical Duty. Capt. Wm. G. McNeill, United States Topographical Engineers, Georgetown, D. C. Georgetown, March 20. 1828. To Capt. W. G. McNeill : Sir ; By the arrangements of the duties of the field for the season, the party destined to examine and locate a line of canal or railway from the mouth of Howard’s Creek, down the Greenbrier and Kanawha Rivers, to the foot of the Great Falls of Kanawha, having been placed under my direction. I have now the honor of submitting a report upon the operations which occupied me during the summer. Early in the season the leveling and survey commenced at the mouth of Howard’s Creek, and was continued, without intermission, down Greenbrier River to its inter- section with Kanawha or New River, and from th nee to the foot of the Great Falls. Returning to the mouth of Greenbrier River, a line of canal was examined from thence to the mouth of Little River. This is regarded as a portion of the Roanoke and Kanawha Canal. In describing the line as located down Greenbrier and Kanawha Rivers, three divis- ions are made : 1st. From the mouth of Howard’s Creek to the mouth of Greenbrier River. 2d. From the mouth of Greenbrier River to Bowyer’s Ferry, (on Kanawha.) 3d. From Bowyer’s Ferry to the foot of the Great Falls. 788 REPORT OF THE CHIEF OF ENGINEERS. FIRST SUBDIVISION. Greenbrier River presents those characteristics which arc peculiar to streams havinoj their i)rincipal sources in the elevated regions of the Alleghany Mountains. In pursu- ing and torcing its circuitous passage thiongh the many ridges of mount.ains that are intersected, it is frequently contined within very narrow limits, and in such cases t is invariably bounded by steep, ragged, and often piecipitons banks. At other points the liills gradually recede from the river, and rich Ha s appear, which offer every facil- ity for the construction of works < f art. To obtaui iirnutely every feat ure of the valley, and thereby insure a judicious selec- tion of that side along which a line of canal might be most advantageously l-'cated, it was thought expedient to examine carefully both banks of the river. From the facts thus a curat ly develo{)ed, it will be seeo that the right bank is decidedly to be preferred, both from its features being generally more favorable, and from its southern exposure. in many irstance=, where bluffs are met with on one bank, more favorable ground might be found on the other. But as the river is remarkably serpentine in its course and cliffs occur at every bend — generally on the concave tide — little or no advantage in any one instance would be gained by cr* ssing. The first level of the first subdivision commences on the right bank of the river at the point (a) on the map, 16 feet above low-water mark. The space, 350 yards, included be tween this point and the termination on the left bank of the canal-line down Howard’s Creek, is occupied by two locks, an aqueduct, and a small portion of canal. The highest freshets on Greenbrier average from 9 to 13 feet; in one or two instances it has been known to exceed 15 feet, but this rarely occurs. In order that the aqueduct may be placed entirely beyond the reach of the highest freshets, and the quantity of drift-wood and ice which is brought down during the w’inter, the water-line is supi)osed 33 feet above low-water mark. The length of the aqueduct is 167 yards. To descend from this level to the level of the first section, two locks of 8^ feet lift are required. A supjdy of water taken from the Greenbrier River at a suitable point above may be received at the commencement of the first level under the mest favorable circum- stances. in order to avoid a tedious repetition of the same facts, which continually recur, a table is attached, exhibiting the length of each level, the number of culverts required, the extent and h--ight of walls, and the general character of the ground. As far as it has been piacticable, the line of canal has been locat'-d at the foot of the hill-Gope, about 15 feet above the surface of the water. This position is recommended by the greater uniformity which it offers m excavation atd the greater facility with which any given level may be retained. From the first to the end of the fifth level no material difficulty is met with. The line is traced alternately through flats, generally narrow, but wifle enough for all prac- tical purposes, and along the river-bank, often rocky and steep. The streams which are crossed are generally small mountain-drains, which yield during the summer months little or no water. The quantity, however, of stone and gravel which is brought down by the current after heavy rains renders it advisable, in most cases, to pass them under the canal through small culverts. At the beginn'Dg of the sixth level the line intersects rugged cliffs of limestone, to sustain the level of the canal, along wdiich a wall 863 yards in length and 20 feet high is required. Its construction is facilitated by the suitalleness of the stone on the sjtot and the ease with which its base may be placed below low-water mark. The water along the right bank, though rapid, is generally shallow. From i hence to the end of the seventh level the ground is quite favorable. In the eighth level perpendicular clifft of limestone again occur. The water at the base of these cliffs is deep and rapid. Near the end of this level the line crosses Mill Creek by means of an aqueduct IdO feet in length and 17 feet above the water. This stream is remarkable from the singular fact of its disappearing entirely 9 or 10 miles from its mouth, and, after flowing in a subterraneous channel through this distance, it suddenly gushes up in a ravine at the foot of high perpendicular cliff-, and soon after puts in operation a mill, which is kept in motion by it throughout the driest seasons. A short dist trice below the mouth of Mill Creek this level is again embarrassed by cliffs of limestone, requiring a wall. Opposite this point a long rapid commences, and a suitable point may he found for a dam ; one s or 9 feet high, and a feeder 946 yards in length, would furni.-h a fresh supplj^ of water at the commencement of the tenth level. Early in this level vertical cliffs of limestone are met with, which will require a strong wall, having its base generally beneath low-rvater mark ; water along the base deep and rax>id. APPENDIX V. 789 From thence for some distance the j^round is favmrahle. Toward tlie end of tlio level tlie line intersects Muddy Creek Mountain, alon^ the declivity of which the canal necessarily occupies very unfavorable f^round. The general slo|)e averages 18°, and is covered for the most part with fragments of sandstone and conglomerate rocks. About the beginning of the fourteenth level tlie mountain recedes from the river, and the lino occupies a very favorable flat until it inteisect'* Muddy Creek. This creek is crossed by a small aqueduct 50 feid. in length and 19 feet above the water. The quantity of water which Muddy Creek yields during the summer months is quite small. Should its supply, however, ever be required, it might easily be thrown into the canal by a moderate dam and a short feeder. From hence, through the sixteenth, imrt of the seventeenth, eighteenth, and part of the nine eenth level the line is traced along a hill-sloiie, frecpieutly steep and rocky. Several culverts are required and some walling, tlie extent of which is shown in the table. I'hrough the remainder of the niueteeth level the nature of the soil admits of easy excavation. Through the twentieth and part of the twenty-first level the line occupies a very unfavorable position along a steeji and rocky hill-slope. From B. M. No. 20 to the end of the last level, with the exception of a straight wall along cliffs of red sandstone, the ground is generally quite favorable. In the twenty-second level, Hunger’s Creek is crossed by an aqueduct 50 feet in length and 17..5 feet in height, with an embankment at either extreniitv of 50 feet. This stream discharges a quantity of water in wet seasons, but during the summer months it nearly goes dry. The deep cutting along thi.s level is very considerable, owing to the general steepness of the hill-slope along which it is located. Toward the end of this level the river suddenly turns from a southeasterly to a northwesterly direction, forming thereby Strieker’s Neck.- On the east side of this bend, cliffs of compact sandstone rise vertically from the water. This point presents a greater obstacle thau qe have yet encountered. The most unfavorable portion extends 266 yards; in this distance the river descends 10 feet, and the whole courseof an impetuous current is concentrated at this point. The base of the wall required here must una- voidably be placed beueathed low-water mark, in deep and rapid water, and should be formed of the strougest materials, to resist the powerful pressure to which it will be exposed during a fresln-t. No opportunity is offei’ed of avoiding this difficulty by crossing the river, as cliffs of an equally unfavorable nature present themselves on the left bank. Thepr file is the development of the linearouml this point. The distance, however, may be lessened more than half a mile, and th s obstacle avoided, by tunneling 136 yards through this neck, or by a deep cut 76 feet. The point alluded to is contained between the letters x and y on the map and profile. Through the next five and part of the sixth levels the canal must also be supported in a great measure by a strong wall, having its base frequently beneath low-water mark. From the middle of the twenty-eighth to the thirty-fifth level the ground is gener- ally favoiable, with the exception of one or two points, at which the hill-.'-lope be- comes steep and rocky. In this distance numerous small streams are crossed, which will require, for the reasons given above, small culverts. About the middle of the thirty-fifth level rugged cliff's of sandstone are encountered. The river fere is wide and shallow, with little fall. About the beginning of the thirty- sixth level, the line gradually curves along the base of the hill, and finally gains the wide valley of Kanawha River. The first suhdi vision is supposed to end at B. M. No. 28, near the commencement of the thirty-sixth level, and opposite the mouth of Green- brier River. Length of the first division, 48 miles 1,561 yards ; fall in the river, 287.52 feet. SECOND SUI3DIVISION. The valley of Kanawha River presents, in a great measure, the same characteristics as that of Greenbrier River. Piobably tliere is not another stream in the country, of equal size, which furnishes less bottom-land along i s banks, and which receives so few tributary streams of any size. The flats which are found are generally so narrow and occur so seldom that the valley is, comparatively, a perfect wilderness. The nu- merous rapids and falls, which are met in nearly every mile, form a striking feature of this river. The bed of the stream beii g very wide, and the water allowed in most cases to flow off’ freely, the freshets seldom exceed 13 feet. Although the soil of the flats is generally light and sandy, the hill-slopes are usually covered with a thick growth of fine timber, such as the poplar, the beech, maple, white and red oak, and some pine trees. The numbers of the levels are supposed to continue from the commencement of the first division. From the commencement, then, of this division to the end of the thirty-seventh level the line passes through a sandy flat, offering but little impediment. Through 790 REPORT OF THE CHIEF OF ENGINEERS. the next two levels, the declivity of the hill being steep and rocky, some emba^ kinents and walling are required. The extent and height of each is exhibited in the table. Opposite The last level the stream descend--, in a few yards, 13 feet, forming what is called “Tug Falls.” The river at the head of this fall is rather wide; in every other respect it presents a veiy favorabl-i site for a dam. From hence, throngn the next tive levels, the line meets with no obitmctions. In the forty-second level Brooks’s Falls is passed. The river is here 29s yards in width, and descends 9 feet over a ledge of rocks extending Jrom bank to bank. This is a more favorable point for furnishing the canal with a new supply of water than the one notic-d above. A dam 8 or 9 feer high, and a feeder less than half a m le in If-ngth, would effect this object. In the forty-fitth level the line intersects the cliffs just above Richmond’s Falls. Tnis fall is a very pronruenC fearnre of the river, and the obstacles here, on the ri>ht bank, are of the most serious nature. The river descends perpendicularly over a ledge of rocks 23 feet high, and is bounded on either side by high vertical cliff’s of compact sandstone. On the right bank the cliff’s commence some yards above the falls, and extend thiongh the forty-sixth and part of the forty-seventh level ; in all which distance the bed of the canal must necessar ly be upheld by a high and strong wall. On the left bank the same rude cliff’s extend for some distance above. Immediatelj’^, however, at the fails a small flat occurs, which offers greater facilities for the construc- tion of locks, &c. At the termination of this flat the bluffs again present their hideous front, and the left bank then becomes as unfavorable as the opposite one. Hence, the idea of crossing the river to avoid the great difficulty immediately at the falls is pre- cluded by the simple fact that the obstacles on the left bank, both above and below, are equally great. Through the remainder of the forty-seventh level the line occupies a narrow, sandy flat. Ftir the next mile and a half, in the forty-eighth level, the hill- slope is steep and rocky ; this is again succeeded by a favorable flat, which continues to the end. In crossing the valley of Meadow River, the embankment is greatly in- creased, owing to the necessity of retaining the level high enough to place the canal above the reach of the high freshets to which this small stream is subjected, as indi- cated by the quantity of crift-wood collected on the flat at its mouth. Aqueduct, 33 j-ards ill length and 14 feet high. Through the next nine levels the canal is generally located along sloping ground, varying from two to twenty-throe degrees. In two in- stances some embankment and walling are required. In the fifty-ninth level l.aurel Creek is passed. About a quarter of a mile above the mouth this creek forks, and discharges its water into the Kanawha through three dif- ferent channels. A small dam at the forks would throw the water into one channel and save the expense ot two culverts. This level, after passing through a sandy flat, encounters a very rugged hill-slope, along which a strong wall must be constructed. Owing to the sudden bend wnich the river here makes, the right bank is exposed to the full force, during high tides, of an impetuous current. Through the remainder of this level and the next the ground is more favorable, sloping from one to 10°. From hence to the end of the seventy-first level the canal is locatetl, for the most part, along a steep and rocky hill-slope. The degrees of declivity, given in the profile, along each level will show in each case the extent of deep cutting. The table will also exhibit the length and height of each portion of walling. Opposite the end of the sixty-lifth level the river descends 11 feet in a few yards, and a favorable site may be found fora dam. To receive this fresh supply of water, the level of the canal is lowered by two locks, with a basin intervening, of 171 yards. Owing to the bed of the stream being here very much contracted, the height of freshets is unusually great. The wall, there- fore, required to sujiport the bed of the canal along this portion of the line must beat least 27 feet high, and calculated to resist the pressure of the immense body of water which rushes by during high tides. From the seventy-second level to Bowyer’s Ferry the ground is generally unfavor- able. In a few instances the line passes through narrow flats, but generally along the hill-slope, which is frequently steep and rocky. Near the beginning of the eighry- fourth level a suitable site occurs for the erection of a dam. A fresh supply of water is thought of thus early, from the circumstances that no convenient site for a dam is found between this point and the end of the line ; and the canal could not be lowered below this without materially jeopardizing its safety. The second subdivision is sup- posed to end at Bowyer’s Ferry. Length of this division, 45 miles 901 yards. Total fall in the river, 399.415 feet. THIRD SUBDIVISION. From Bowyer’s Ferry to within 2 miles of the Great Falls the valley of Kanawha presents a novel and frightful appearance, and the obstacles that are met with are decidedly more formidable than any that have even yet been encountered and over- come The river here breaks through Sewell Mountain, and is confined throughout within narrow limits, bounded Iiy rugged banks and jirecipitous cliffs, rising many hundred feet above the surface of the water. Immense masses of rocks, which have been ejected from the cliffs near the brow of APPENDIX V. 791 the mountain, cover the lower slope on both sides of the valley, giving to the whole a most appalling? appearance. In many instances these hn^e fragments have been precipitated entirely across the bed of the stream. In such ])Iace8 the river when agitated and swollen by heavy rains — the cnrrent being obstructed in its course — fre- quently rises from 35 to 50 feet above its usual height. Numerous rapids-and falls occur, over which the water rushes with deafening impet- uosity. From this faint account of the general character of this portion of the route, it will immediately be seen that the bi-d of a canal must unavoidably be supported tf.rough- out the whole distance by a high wall, capable of resisting the immense pressure to which it must necessarily be exposed from the impinging of this furious current. There is generally sufficient space (with the exception of threeor four points noticed in the profiles, when the cliffs i ise vertically from the surface of the stream) between low-water mark and the blufis for the construction of such a wall. This space, how- ever, is usually occupied by the immense masses of rocks alluded to above. In all this distance there is but one small flat; hence, the want of materials on the spot for puddling the bottom and sides of the canal will be very seriously feP.. In making the survey every feature of the valley was so carefully observed that, in case a canal along this portion of the route should be thought inexpedient, from the unusnally great expense which would attend its construction, a railroad might be sub- stituted a ong the same line, with such alterations as might be suggested by the data now in our possession. This, however, must hereafter form the subject of a separate report. About half a mile above the month of Gauley River the line gains a narrow flat, along which it continues until it crosses the river by an aqueduct 23 feet above the surface of the stream and 213 yards in length. From hence to the Great Falls of Kanawh i the line o'^cupies a favorable flat. To descend from the level of the last s ctiou to the surface of the water at the foot of the Great Falls four locks ot 9.8 feet lift are required. A ledge of rocks, extending diag. nally across the river and over which the stream descends perpendicularly 21 feet, forms what is known as the Great Falls of Kanawha. The main body of water passes through a sluice about 100 yards in width near the left bank. The remainder of theled^e is only covered at high tide. Just below the falls there is a most beautiful natural harbor, the river being here 650 yards in width and 10 to 15 feet deep within a few yards of the shore. Length of third division, 21 miles 1,638 yards ; total fall in the river, 340.325 feet. In conclusion, it may be proper to remark that hereafter, on making a more minute and protracted examination of the ground preparatory to a permanent location, many alterations may be suggested in the position of the line of canal and its appurte- nances. But it is believed that no material difference in the total expense of the work will result from such partial deviation. All which is respectfully submitted. R. A. Hazzard, Lieutenant, United States Army. 1 Number of levels. j Length of each level. Number of locks to the beginning of each level. Number of aqueducts in each level. Number of culverts in each level. Walling. Nature of excavation. Yards. Length. Height. Soil. Hocks. Feet. Feet. 1 0 1, 466 1 Clay Sandstone. 2 2 3 3 Sandy 3 2 1, 193 4 2 Clay and sand 4 2 i^OO 5 5 5 1 1, 500 6 2 Clay 6 1 1, 246 7 863 20 Limestone and conglomerate rock. 7 1 170 8 1 Alluvial 8 1 1, 350 9 1 1 255 23 Clay Limestone. (A small aqueduct across Mill Creek. 100 feet ia length and 17 feet high.) 9 0 763 10 Gravelly 10 2 241 11 1 666 18. 5 Stony Sandstone. 11 1 340 12 1 228 10 12 0 1, 438 13 1 13 0 1,233 14 Alluvial 14 1 1, 366 15 1 Clay Valley of this small 8‘ream is 60 feet in width and 13 feet in depth. 7 ( oS "3 ,2 Vi o u a> ,0 a D 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 Number of aqueducts iu each level. OF THE CHIEF OF ENGINEERS. aj . 5.2 Vi _ SI 2 a s ■Wallin er. Feet. 756 810 728 408 776 491 1,077 1,286 850 833 1,266 528 '566 291 1, 347 300 476 925 1,213 ‘"756’ 1, 383 "821' 475 Feet. 17.5 22 23 17.5 20 17.5 25 Soil. Alluvial Ciay Alluvial . Gravelly. Clay Gravelly. Clay Sandy . . . Sandy . . Alluvial Sandy . . Clay . Sand. Clay . . Sandy Clay , Clay Sandy Sandy Clay Nature of excavation. Rocks. Aqueduct across Muddy Creek, 50 feet in length and 19 feet above surface of the water. Slatestone. Embankment 50 yards in length and 15 feet high. Limestone. Hard sandstone. Hard sandstone. Rocky. Sandstone. Sandstone. Sandstone. Compact sandstone. (Two con- tiguous locks are required at Richmond’s Falls ) Compact sandstone. Conglomerate rock. Hill slope, very steep and rocky. Sandstone. Stony. Rocky. Sandstone. Red sandstoneand limestone cliffs Sandstone cliffs. Rocky. Along a very steep hill-slope. Clay Rocky. APPENDIX V. 793 From hence there are in all 42 levels and 46 locks to the surface of the water in the basin below thn Great Falls; but the length of each level below Bowyer’s Ferry, with such other details as are iucladed in the tables, are reserved until the drawings shall be more advanced, on which every featii e of the section will be fully exhibited. The precise termination of each level, with the most suitable position for the locks, in- volves considerations which are not at this time fully niatuied. Georgetown, D. C., Topo(jra])Mcal Office, March 14, 1828. To Capt. Wm. G. McNeill: Sir : I have the honor to submit the following table, exhibiting the discharge of the sevoiat 81 reams which may be U8ed for the supp y of the proposed Roanoke and Kan- awha Canal, from the mouth of Little River to the forks of the Roanoke, together with some few remarks upon those streams and the more important mineral localities ob- served in the vicinity: Table of measurements. Name of stream. Date of meas- urement. Quantity in cubic feet discharged per second. Remarks. 1827. South Fork of Elliot’s Creek, at the forks August 7 5. 48 Do 20 9. 00 Slight fresh. Do 2-2 5. 08 Do October 12 5. 45 North Fork of Elliot’s Creek, at the forks August 7 3. 80 Do 20 8. 00 Slight fresh. Do 22 4. 60 Do October 12 4. 03 E’liot’s Creek, monlh 22 12. 02 South Fork of Roauoke, above Elliot’s Creek 22 28. 00 About the minimum. Meadow Creek, near Hay’s September 21 6 44 About the minimum. Little River, near the mouth of Meadow Creek... 17 231. 04 Little River, 3 miles below Thompson’s G 161. 70 Slight fresh. Little River, at the commencement of feeder to the summit-level 19 118.00 Do October 4 97. 00 Do 14 111.05 The number of mills upon th<=se streams renders complete accuracy unattainable. The measurements were made under the most favorable circuni'tances for Laming the miuitnum supply of the strea os; for not only the quantity of rain which fell dur- ing the season was unusually stnall, but the intervals of its falling were so great as to give ns an opportunity of ob-erviug the strength of the spr ngs. Little River, I iiuifurmly learned from tho^e who resid d on its banks, had never been low'er than during this season, so that its minimum supply may, with common certainty, be assumed at what it is represented in the table. In comparison with otherstreams in the vicinity ^tfording the same quantity of water, Little Rivt-r is very much the shortest, and by far the greater proportion of the supply is obtained from springs which have their rise wirhiu a few hundred yards of the river, a distance general y cleared along the whole of its length. No usual occurrence, there- fore, can tend to lessen its supply materially ; but should any circumstances whatever produce this effect, we may have resort to a reservo r, for whi- h the valley of the river is well adapted; and a reservoir may also be formed in the valley of the South Fork of Elliot’s Creek, near the fork, for a supply at that point. The count' y is generally what is termed a limestone country, except about the upper part of Litile River, where we have sa' dstoue. Compact limestone is the most abuudaun mineral production , but as to its presence united with the oxide of manganese, a union suitable for water-lime, I had not the means of ascertaining. A fine locality of bubrstone was observed on the South Fork of Elliot’s Creek, 12 miles f om the Roanoke, containing a great proporti'on of silex, and very porous. Iron ore was nonced on Little River, ferruginous red oxide of copper on Elliot’s Creek, and galena or sulphuret of lead on New River. I am, sir, very respectfully, your obedient servant, John M. Fessenden, Second Lieutenant Fourth Artillery, on topographical duty. 794 REPORT OF THE CHIEF OF ENGINEERS. Topographical Office, Georgetown, JJ. C., February, 1828. To Capt. William G. McNeill, United jStales Topographical Engineer : Sir; Agreeably to your instructious, I have the honor to report upon the duties as- signed to me while employed upon the surveys and examinations relating to a pro- posed route for a canal between the headwaters of the Kanawha and James Rivers, in the State of Virginia. During the months of August and September, 1826, and the months of May and June, 1827, I ascertained, by frequent measurements of t he streams on each side of the dividing-ridge, the quantity of water to be relied upon for the supply of the summit- level, together with ihe portions of canal between the extremities of the summit-level and .Jackson’^s and Greenbrier Rivers. In the mean time I also made an examination as to the supply of water for a canal to connect the headwaters of Craig’s Creek, a tributary to Jackson’s River, with Sink- ing Creek, a branch of the Kanawha. During the latter pait of the season of 1827 I reconnoitered the country in the vi- cinity of Huntersville, Pocahontas County, to ascertain whether or not a connection could be formed there between Jackson’s and Greenbrier Rivers. While in the performance of these duties, agreeably to instructions received from you, I made fiequent observations relative to the formation of the country which I traversed. I shall give the result of my observations in the order in which they were made; showing, in the first place, that there is a sufficient supply of wmter for a canal on the route from the mouth of Dunlap’s Creek, at Covington, to the mouth of How- ard’s Creek. The following tables contain the quantity of w'ater discharged by the several streams to be relied uxion for the supply of a canal on the dilferent routes examined: Table I. — Exhibiting ihe quantity of water obtained by the measurement of streams on the proposed route for the James River and Kanawha Canal, in Virginia, during the months of August and September, 1826. Name of stream measured. When measured. Where measured. Quantity of water per second in cub. feet. Jackson River... Do Do Do Do Dunlap’s Creek .. Do Do Do Do Do Do Snake Run Aufile.v’s Creek . . Potts’s Creek Do Do Craig’s Creek Do John’s Creek Do Do Sinking Creek .. Do Greenbrier River Do Do Do Do Do Do Do Do Do Do Do Do Do Howard’s Creek . Do Aug. 7 At Pitcher’s mill, near Covington 7 Above the mouth of Dunlap’s Creek 24 Above Pitcher’s mill Sept. 6 Above ihe mouth of Potts’s Creek 4 Below the m uth of Dunlap’s Creek Aug. 7 Below the mouth of Augle^’s Creek 7 At its mouth 9 Above the mouth of Snake Run 16 At its mouth 17 .do 24 do Sept. 27 do Aug. 9 do 7 7 do 16 do Sept. 4 do Aug. 25 Below the mouth of Roaring Run Sept. 1 Below the mouth of John’s Creek .. 1 At its mouth 1 do 1 . ... do 1 Below Johnson’s mill 1 At Johnson’s mill, when there is a maximum head of water at the mill Aug. 10 Above the mouth of Howard’s Creek 20 do 20 At Bright’s mill 21 One mile above Bright’s. 25 Above Howard’s Creek Sept. 5 Below Howard’s Creek 6 Six miles below Anthony's Creek 7 At Bowen’s mill 7 Below Bowen’s mill 7 Above McClure's mill 7 At McClute’s mill 8 Above Laurel Run 8 Above Stamping Creek 25 Below Howard’s Creek Aug. 10 Below the White Sulphur Springs 19 do 130.5 157.2 1.52. 05 156.8 246.3 22.6 2.5.6 19. 03 9. 43 10. 26 24.0 22. 0 5.5 1. 12 65.3 34.9 39.6 157.7 39.3 15. 4 16.3 15. 85 0. 49 6. 40 230.7 134.7 69. 44 70. 00 148. 6 97.0 46. 72 46.7 129.0 73.3 67.6 49.0 42. 3 142. 6 11.6 7.8 APPENDIX V. 795 Table I . — Exhibiting the quantity of water obtained, ^c. Name of stream measured. When measured. Where measured. Quantity of water per second in cub. feet. Howcir^l.^9 ... Aug. 25 Sept. 5 5 A tits month 6. 83 Do do 12. 9 Do do 10. 9 {) do. 14. 3 Do Aug. 21 Six miles above its mouth 11. 5 Spring Creek nT'APir . . . Sept. 7 8 Near it-* junction with Greenbrier Kiver Near its mouth 9.6 0. 66 Do 8 At a mill near its mouth 0. 704 Sf.jimpi HIT rirop.k 9 Near its mouth 5. 00 9 do 2. 5 Knapp’s Creek 14 Six miles above its mouth 41. 0 ^pfuiTifi . . Aug. 29 29 Below Patton’s mill 33. 3 Do At Knox’s Ford 28. 29 Do 29 Near the Gap Mill 30. 09 Table II.^ — Showing the quantity of water found in the streams on the proposed route for the James River and Kanawha Canal, in Virginia, during the mouths of May and June, 1827. Marne of stream measured. When measured. Where measured. Quantity of water per .second in cub. feet. Jackson’s Kiver May 4 Below the mouth of Dunlap’s Creek 3.50. 00 Do 18 56"*. 78 Do ' 19 do - .. 607. 00 Do 25 do 1, 800. 00 600. (-0 )o June 5 16 May 9 10 do - Do do ... 419. 70 Green bi ier Kiver Below mouth of Howard’s Creek 1,768. 00 1, 995. 00 1, 970. 60 4, 000. 00 Do do. . - Do 16 do Do 12 do - - Do 22 do 6, 650. 00 2, 000 00 1, 500. 00 1, 224. 00 Do 24 do Do 25 do Do 27 A t Bowen’.s mill ... Do 27 At McClure’s mill 8f<2 00 Do June 5 A t the bridge 1,800. 00 752. 30 Do 26 Near McClure’s .. Augley’s (or Ogley’s) Creek.. May 1 At Callaghan’s mill 10. 80 bo June 16 At its mouth 9. 50 Do May 21 June 26 do 50. 00 Do do 60 00 Howard’s Creek May 16 21 do - - _ . - . . _ 40. 40 Do do 370. 00 Do 21 Below Dickson 's Run . 49. 30 Do 21 Near its mouth . . ... 184 60 Do 25 do 180. 00 Do June 6 do . .. . 360. 00 Do 7 Near Little D ckson's 67. 60 South Fork of Howard’s Creek May 16 Near the toll-gate 5. 57 Do 22 do 80. 00 Do 24 25 do 30. 00 Do do __ 26. 00 Do June 19 do 27. 00 Dunlap’s Creek May 17 17 Above Colonel Crow’s . 41. 50 Do Below Crow’s mill 52. 90 Do 19 21 At its mouth 79. 30 Do Below Sqnii’e Bishop’s . .. 323. 02 Do June .5 Below A ugley’.s tb’eek 50. 00 Do 16 do 19. 80 Potts’s Creek May 19 23 At its mouth 8.5. 10 Mill Creek do 142. 72 Dickson’s Kun 24 Near its junction with Howard's Creek 31. 00 Spring Creek 27 At Burr’s mill .... 56. 00 Locust Creek 27 Near its junction with Greenbrier Riv'er 84. 00. Anthony’s Creek J une 27 One mile below Wiley’s, and near the intersec- tion of the feeder -line, cC-c. 433. 10 Note. — A ll the streams ruentioned ia the f.tres'dng tables were measured by means of a float, with the exception of Dunlap’s Creek, which, on the 10th and ITtb days of Aujiust, was measured by means of a waste-wier or dam. At mills, the method prescribed for measuring by means of orifices lias geu- erally been adopted. The quantity of water given in the fourth culurans of the tables is expressed in cubic feet per second. 796 REPORT OF THE CHIEF OF ENGINEERS. Table III. — Of the qnatitiiy of water found in the streams on a proposed route for the James JUrer and Kanawha Canal, in the vicinity of Huntersville, in the county of rocuhonias, Virginia, (May and June, 18‘27.) Name of stream measured. When measured. Where measured. Discharge of water in cubic feet. Greenbrier River Do May ‘29 June 2 Below the mouth of Knapp’s Creek Seven miles above Ihe mouth of Deer Creek 2, 400 420 Do 10 Ten miles below mouth of Knapp's Creek 432 West Fork Greenbrier River 2.5 At its iunction with East Fork Gieenbrier River 777 East Fork Gieenbrier River . . 25 At its iunction with West Fork Greenbrier Rivei 220 Knapp’s Creek May 29 :u Near Huntersville 156. 9 Do Above the mouth of Sugar-Tree Creek 1.5.0 Do June 10 78 Warm Spring Run May Four miles below Warm Springs 28 Do •June 9 do 14 Jackson’s River May 30 Midway between tbe Warm Springs and the 448 Do J une 9 month of Little Back Creek. do 140 Big Back Creek May 30 Above the mouth of Little Back Creek 296 Do . . •June ic do - 140 Little Back Creek May 31 'une 10 At its mouth 50. 6 Do ... do 20 Sugar-Tree Creek May 31 Near its mouth 15 Sitlington’s Creek 31 Neat— wberi^ a feeder-line would strike it, &c 100 Deer Creek June 1 Above Woodle’s mill 86 Do 19 ... do 50 Camp Run, a branch of Deer 1 At its mouth 11 Creek. Duncan’s Run, a branch of 1 do 10 Deer Creek. Buffalo Run, a branch of Deer 1 do 10 Creek. Salisbury Run, a branch of 1 do 11 Deer Creek. Back Cieek 24 Near McCloud’s 200 Knajip’s Creek 19 Above the month of Sugar-Tree Creek 10 Six other small runs 19 Where a feeder would strike them, &c 16. 00 I will now point out the resources which may he relied upon to supply with water the snnunit-level and the portions of canal on the Covin^rou route between the mouth of Dunlap's Creek and Howard’s Creek. It is known Irom the surveys which have been made on this route thnt we must, in a measure, depend for a supply of water upon the following streams, viz : Cub. ft. Greenbrier River, (minimum supply,) 8th September, 1826 42.3 Greenbrier River, (mean supply,) 1826 67.6 Anthony’s Creek, 6th September, 1826 11.6 Dunlap’s Creek, 9th September, 1826 5. 5 Besides the supply from these streams, we can avail ourselves of a reservoir, which shull contain either 13,060,584.9 cubic yards or 4,603,040 cubic yards, according to the height of the dam which may be constructed at the point where the feeder from Green- brier River is taken. To show that this supply will suffice to feed the summit-level, to supply its lockage, and feed, besides, the portion of the canal contiguous to it, w e have made the same suppositions with regard to this route that ihe board for internal im- provement did with regard to the proposed route for the Chesapeake and Ohio Canal. Suppose, in the fiist place, the navigation to be interrupted four months in ihe year, viz, from the 1st of December to the 1st of A{)ril. Adopting 67,6 cubic feet per second as the mean supply of water afforded by Greenbrier River hist, in the winter season, at the point w’here the feeder to the summit cou.mences in said stream, it w ill be found < that the larger reservoir would be filletl twice during the iufeiruption of the naviga- tion, and the lesser leservo r w< uld be tilled in nearly one-thiid of the time required to till the larger one. But 67.6 cubic feet per second is certainly much less than the mean supply afforded by the Greenbrier River in the winter season. (See Tables II and III.) It is, in fact, thought to be much less than the mean supply during the sea- son of open navigation. So that, w hile the navigation continues open, upon this last supposition the larger reservoir would bo filled at least every two monttis, and the smaller reservoir oftener than opce per month from Greenbi ier River itself. Now', in order that we may see what influence evaporation, &c., may have ue on this supply, let us make a computa- tion as to the supply which may be expected from rains. This, it will be seen, besides making up for evaporation and soakages in the reservoir, i APPENDIX V. 797 and in the feeder to the smnmit-level, will leave a very large snpplj’- which may go towHid the feeding of the snrnniit and portions of the canal contignons to it. We shall take an area of 55 sqnaie miles, which we su[)pose to supply Greenbrier Kiver with rain-wuiter. Tins area is chosen fioni an inspection of the map of this jiart of the conirry, from which we find that the sum of the lengths of the streams trihntary to Greenbrier River exceeds 110 miles ; and, frosn a knowledge of the country, we know the width of tiieir valleys to average niore than half Ji mile. We shall assume for the fall of rain on this surface the least quantity w'hich fell from the year 1817 to 1824, in the vicinity of Baltimore. Tlie least quantity fell in 1822, whicli was in the fall and winter, 16.7 inches, and for the six other months, 12.5 inches; making, for the \vhole year, 29.2 inches. So it will he found that the assumed area will receive during — Cubic yards. The fall and winter 78, hAo, 3h4 The sjuing and summer ' 59. ll~,696 The whole year round 137, 998, OdO From which it will he seen that the first quantity would he six times as much as would he necessary to fill the larger reservoir in the winter season; and seventeen times as much as would fill the smaller one; and that the second quantity would till the larger reservoir more than four times in the summer, and the lesser one nearly thirteen times. Now, what is the evapcu’ation, soakage, &c., compaied with these quantities? Suppose the evaporation to be 32 inches [ler annum, which are equivalent on a surface equal to 1 sqna e yard to 0.910 cubic yard. We will apply this to the feeder and reservoir. The feeder, w'hich is about 33f miles in length, will have a sur- face, say, equal to about 590,880 square yards ; and, consequently, the evayior.ition on it would be equal to 541,300.8 cnbii; yards. Suppose now the soakage to be equal to one and a half times the evaporation, (w'hich is, no doubt, considering the natnie of the soil over which the feeder-line passes, a snffici'ent allowance.) This last sn))position gives for the evaporation and soakage together, of the feeder, 1,353,252 cubic yards. The evaporation of the reservoir, taking the larger one, whose surface would be ecjnal to 2,508,333.3 square yards, wonhl be equal to 2,282,583,303 cubic yards ; or the i-vapora- tion and soakage together equal to 5,706,458.25 cubic yards; making, for both the feeder and reservoir, the evapor^ition and siaikage together equal to 6,059,710.25 cubic 3 'ards. Should we double the allow'auce above made for evaporation, &c., it would be seen that it could not even then be compared with the quantity available from rains. It is, in fact, very trifling, even when compared \\\ h the fall of rain dut iug the s|)i iug and summer months. It should also be recollected that we have assumed for the fall of rain the least quantity which fell in the course of a number of years in a section of the country where the fall is not to be compared w'ith that in the elevated section of countiy which we have examined. I think, therefore, that it would be safe to adopt the following as the minimum supply of water w'hich can ever be expected on the route in question, and w hich will suffice to feed the summit-level and portions of canal contiguous to it, viz: Cubic feet per second. Greenbrier River 42.3 Anthony’s Ci eek 5. 5 Dunlap’s Creek «... 5. 5 and the supply of 13,060,584.9 cubic yards from the reservoir, (using the larger one.) It will be observed by reference to Table I that the least quantity of w'ater found in Anthony’s Creek, 6 miles from its mouth and near where the feeder-liile strikes it, is 11.6 cubic feet per second ; we have assumed less than half of it for the minimum supply. A less quantity is also assumed for Dunlap’s Creek than has ever yet been found in it at any one point; one of its tributaries. Snake Run, gave, on the 9rh of August, 1826, the same quantity which it will be perceiv'ed is assumed for the stivam itself. And it should be recollected that the canal-line strikes the creek some distance below the mouth of Snake Run. We could no doubt avail ourselves of several time.3 the quantities assumed as the discharge of these streams; for, besides the actual quantities flowing in them, we can, if necessary, have recourse to small reservoirs on both of them ; and, as reganls Green- brier River, it presents frequent sites for dams suitable for the formation of reservoirs ; and that, should it be found necessary, one may, of almost any height, be formed at any point above the one chosen for the commencement of the feeder to the summit-level. If necessary-, reservoirs might be formed upon several of the tributary s. reams of Greenbrier River, and sometimes upon streams tributary to them. But to Hunt the supply to the quaniity above assumed, and calculating upon an open navigation of eight months in the y'ear, it w'ill be found that the monthly supply (adopting the larger reservoir) will be as follows, viz: 1,632,573.1 cubic yards from the 798 EEPORT OF THE CHIEF OF ENGINEERS. reservoir, 4,ir>6,790.4 cubic yards from the Greenbrier River, and 1,055,980.8 cubic yards from Dunlap’s and Anthony’s Creeks, making the whole supply equal to 6,845,344.3 cubic yards. Tills monfhly supply may be disposed of after the'following manner ; Considering tlie length of the summit-level, it is probable that about two hours would be required for a boat to pass from one extreme point of it to the other — that is to sav, it would move at the rate of a little more than one mile a id a half per hour. This would be for a singlo boat. But, with a view to the saving both time and water, we will sup- pose the trade to be carried on by a number of boats, passing and repassing the differ- ent locks at the same time. Let the number be twenty. But, as a greater length of time is required for a number or train of boats to pass the summit-level than for a single one, we shall increase the time assumed for the passag^i of a single one one-half for a train — that is, we suppose it to take three hours for a train of 20 boats to pass from one end of the summit-level to the other. ,The trade from the East to the West, we suppose, would be carried on in this man- ner. The ti st tniu of twenty boats would leave the first lock at the commencement of the summit-level, and arrive in three hours afterward at the 1 mk at the other ex- tremity of the level, (which we shall here call the second lock,) when it would meet a second train, having jiassed the second lock. This second train would arrive in three hours afterward at the first lock and find a third train, having passed duiing the pas- sage of the first and second trains, and ready to proceed toward the second lock, and vh’ch would arrive there in thr<-e hours afterward, and find a fourth train Imving ascended t the summit and ready to proceed to the first lock ; and so on, until the whole five trains shall have passed the summit. The passage of these five trains, or one hundred boats, requiring fifteen hours, may be assumed as the greatest trade on this part of the canal. It is readily seen that this trade agrees very well with the monthly supply, for, at the rate of 100 boats per day, 3.000 might pass per mouth, and 24 000 during the sea- son of open navigation ; and as the boats are supposed to pass the locks alternately, one lock full of water onlj’ for each boat will be required to pass the summit level. But to remove all fioubts as to the adequacy of the monthly supply, we shall add half a lock-full for leakage, &c. This allowance, supposing the size of the locks the same as those contemplated by the board of int rnal inqirovements for the proposed Chesa- Xieake and Ohio Canal, will be equal to 623 cubic yards; so that all the boats which W'ould pass during a month would require for lockage 1,869,000 cubm yards of water, which, when taken from the monthly supply, leaves 4,976,344.3 cubic yards. This residue would supply the canal from the mouth of Howard’s Creek to the mouth of Dunlap’s Creek, a distance of about 33 miles 737 yards, at the rate of 148,992.:l4 cubic yards per mile per month, or 93,120.2 cubic feet per miuute, which are equal to 1,552 cubic feet per second per mile. And for the supply of the whole distance from the mouth of Howard’s Creek to the mouth of Dunlap’s Creek, we sh >uld have considerably more than 1| cubic feet per second fier mile, should we take into consideration the quantity which may be had from Howard’s Creek, some distance above its month. We speak now of ^ cubic feet of water per second, exclusive of lockage, and, if you please, of evaporation, soakage, &c. It is thimghl., however, to be a sufficiency, absorption, evaporation, &c., included. We might assure ourselves of this fact immediately by making a comparison between this supply and the different supplies which answer for many of the canals constructed in France and elsewhere. Bu 1 shall merely advert to the fact that, from experiments which havb been made upon the New York canal and others which have been con- structed in this country, it has been ascertained that about 100 cubic feet of water per minute per mile oniy are requisite for navigating an ordinary canal in this countiy, lockage included. Now, if we ap])ly this to the James Rtver and Kanawha Canal, on a distance of about 33 miles 737 yards, and deduct 1.869,000 cubic yards, which we have allowed for lock- age, we shall find that there will be a much smaller quantity left (and which would supjily the canal) than we can calculate upon from Greenbrier River, Anthony’s and Dunlap’s Creeks, and the reservoir which would be formed ou the Greenbrier. The foregoing calculations were made from data obtained from the experimental surveys made in 1826. Subsequent examinations develop facts which effect a slight change in them. By the recent surveys which have been made it appears that the feeder- line, from Greenbrier River to t e summit-level, can be shortened about 2 miles, making the entire length of iti o; ly 31 miles 130 yaids. The length of the summit-level is 4 miles 789 yards, and 'his distance from its east- ern extremity to the point where the canal is fed by Dunlap’s Creek is 2 miles 1,718 yards. I he distance from thence to the mouth of the cretk, 16 miles 115 yards. The di-tance from the point where the feeder-line strikes the summit-level to the point where the canal is fed by Howard’s Creek is 3 miles 1,175 yards, and from thence to the mouth of the creek, 4 miles 740 yards, making the length of the canal, from the mouth of Howard’s Creek to the mouth of Dunlap’s Creek, about 31 miles 1,017 yards. APPENDIX V. 799 An alteration lias been made in the situation of the dam for the reservoir on Green- brier River. This dam, thonjijh supposed to be as high as the dam of the larger reser- voir first calculated upon, is found to give a reservoir with a prism of water equal to only 8,57S,14d cubic yards, but which, when compared with the former larger reservoir, is found to have a much smaller than a proportional surface exposed to evaporation. The surface of this reservoir is equal to 532,666 square yards, and the evaporation from it would be equal to 484,726.06 cubic yards. If to this quantity, after making a due allowance for soakag-*, we add the evaporation and soakageof the feeder-line, and subtract the sum thereof from the quantity of water available from rains, the residue would still be more than sufficient to 611 the ado))t(id reservoir once per monUi. Be- sides, it is probable that the reservoir would be filled once per month the year round from Greenbrier River. The quantity (67.6 cubic feet) hrst assumed as the mean sup- ply of the river would 611 it once in le s than every forty days. So, to feed the sum- mit-l 'vel and portions of the canal between its extremities and Dunlap’s and Howard’s Creeks, wm may, with safe>^y, calculate upon the contents of a reservoir formed in Greenbrier River, and the least quantities of water afforded by Greenbrier River, An- thony’s Creek, Howard’s and Dunlap’s Creeks. The supply would be as follows: Calculating upon an open navigation of eight months in the year, the reservoir formed in Greenbrier River would give a monthly supply of 1,072,268.5 cubic yards. And from Gret-nbiier River, containing 42.3 cubic feet per second ; Anthony’s Creek. 5.5 cubic feet per second ; Howard’s Creek, 6.8 cubic fret per second ; and Dunlap’s Creek, 5.5 cubic feer, per second, we should have a monthly supply of 5,769,552.6 cubic yards. If from these supplies we subtract the lockage for 3,000 boats per month, it will be seen that the residue wouhl feed the entire canal from the mouth of Howard’s Creek to the mouth of Dunlap’s Creek at the rate of something like 157,517 cubic yards per mile per month, which are, exclusive of lockage, more than sufficient for that ])urpose, all other losses included. The canal may be fed in this manner : Frohi Greenbrier River the reservoir formed in it, and Anthony’s Creek, aftording togei her a monthly supply of 5,661,068.5 cubic yards per mile. From this quantity there will be left, after feeding the summit-level and the portion of canal between its eastern extretnity and the mouth of Dunlap’s Creek, (a distance of about 23^ miles,) about 618,519 cubic yards p r month, that is, after deducting lockage for both extremities of the summit-level and taking into con- sideration a monthly supply of about 528,000 cubic yards afforded by Dunlap’s Creek more than 16 miles from its mouth. Now, this remaining supply of 618,519 cubic yards per mile, &c , together with a supply of 652,800 cubic yards per mile per momh from Howard’s Creek, more than 4 miles from its mouth, will feed the remaining portion of the canal from the western extremity of the summit-level to the mouth of Howard’s Creek, a distance of about 8 miles. If we found the calculations upon the least quantities of w^ater which have yet been found to flow in the streams U'Cd on this route, at the points where it is prob- able they would be brought into requisition to fe-d the canal, it will be seen that the fall of rain-water need not be so much relied upon. For instance, we have Greenbrier River, containing 46.72 cubic feet per second, (see T.b. 1;) Anthony’s Creek, 11.50 cubic feet per second ; Howard’s Creek, 6.8;i cubic feet per second; and Dunlap’s Creek, 9.43 cubic feet per second ; giving alto ether a monthly supi)ly of 7,150,080 cubic yards. If to this supply we add the monthly sui)ply from the reservoir, which we supposed filled from Greenbrier River during the interrupti »n of the navigation, and make a deduction for lockage, there will remain a monthly supply of 6,353,348.5 cubic yards, which would supply the canal under consideration at the rate of 194,581.2 cubic yards per mile per month, which are, at least, more than half a cubic foot per second per mile more than is necessary for that purpose ; and which, together with a small quan- tity of rain-water, would supply the lo.sses of the feeder, &c. Before coiiclud ng, we will make one other supposition. Suppose an open naviga- tion of nine months in the year; and suppose the fe^^der from Greenbrier River to the summir-level, and the canal from the mouth of Howard’s Creek to the mouth of Dunlap’s Creek, to lose each its prism of water pnr month by evaporation and filtration, and allow 1,454, 178. H cubic yards for the leakage and evaporation of the reservoir, (sup})os- ing the former to be twice the latter,) and making a due allowance for lockage, we shall have, during an open navigation of nine months, all htsses t^qtial to about 32,962,494.18 cubic yards. Now, if we compare this with the fall of raineven during the spring and sum- mer, there will remain a monthly supply of 2,906,133.53 cubic yards of water, which may go toward supiilying the canal. That is to say, we should have lor the supply of the canal, including running water, a monthly supply of 8,771,933.53 cubic yards, after having made good all losses ; which is twice as much as is necessary. I shall finish these remarks by repeating, that, when we take into consideration the many favorable sites for darns, suitable for the formation of reservoirs on Greenbrier River and else- where, we can never dread an insufficiency of water for the supply of a canal on this route. In addition to this fact, the rela ive situation of Cheat River with ivgard to Greenbrier River, by which, in case of any unexpected casualty, a feeder, with a suf- 800 REPORT OF THE CHIEF OF ENGINEERS. ficient supply of water for a canal, inifrlit be brought from thence to the Greenbrier River, leaves no doubt of the practicability of a canal comnmnicatiou between the head- waters of Janies River and the Kanawha. The next nhiect which occu|)ied niy attention was the examination of a proposed route for the James and Kanawha Canal, by way of one of tlie head branches - f Cra g’s Creek, a tributary to Jackson’s River, and Sinking Creek, a tributary to the Kanawha. I deem it nunecessary to enter into a minute descri[)tion of the country along this route, as I shall be able to show in a very few words that there is not a sufficient supjily of water for a canal. To supply the summit-level of a canal on this route we should have to rely almost altogether on John’s Creek, a branch of Craig’s Creek, which gave on the 1st September, lb2(), at i s month, a discharge of water equal to only 10.4 cubic feet per second, which is not perhaps the minimum supply during the driest season of the year. As regards Sinking Creek, we might often during the dry season expect to find it, if not dry, almost so. Relying then principally on John’s Creek, we should have a monthly supply of water which would not Mittice for lockage alone. Allowing the same quantity of water for lockaiJe on this route as we did on theCoving on route, we would want an additional quantity of 391, 5b0 cubic yards per month for lockage. (Referred to on p. 777.) The quantity of water afforded by John’s Creek, (15.4 cubic feet per second,) and the quantity afforded by Sinking Cr-ek, (.49 cubic feet,) together equal to 15.89 cubic feet per second, would have to supply a summit-level and |)ortions of canal, together equal to, at least, from 30 to 36 miles. Now, at the rate (ff 1^ cubic feet per second, it is readi y seen that the above quantity would only feed a canal something like 10 miles, exclusive of lockage. In a word, the great additional supply of water wanted for a canal by this route can never be had, for we can alone expect it trorn reservoirs, and though one or two might be f -rrned on the route they could be but small. A canal by this route, even if there was a sufficient supply of water, would have nothing to recommend it, for we would certainly have a very long tunnel on it, and that, too, over an elevati d summit-level. The foregoing remarks would apply equally as well to any route by the way of Potts’s Creek, &c. The next examination which I made, agreeably to the instructions received from you, was of the country between the head luauches of Jackson’s River and Greenbrier River, in the vicinity of "Huntersville in Pocahontas County. My attenti n was particularly directed to the ascertaining of the fact ‘^whether or not the Greenbrier River, near the mouth of Knapp’s Creek, approached sufficiently near to Jackson’s River or some of its branches ; and whether or not the dividing ridge bei ween those str«*ams was sufficiently depressed to admit of a connection between, them by means of a canal with ut a tunnel.” There are, it is true, several depressions in the Alleghany Mountains, or dividing ridge betw'een the eastern and western waters, in the county of Pocahontas. Tuece is one 7 or 8 miles from Huntersville, between the head- waters of Knapp’s Cre k and Ldtle Back Creek; and another still higher up, between the head-waters of the Sugar Tree Creek, a branch of Knapp’s Creek, and Big Back Creek, a branch of .Jackson’s River. But the most considerable one in the mountain is where the road from Huutersville to Warm Si)riugs, in Bath County, crosses it ; a communication might be made be- tween Knapp’s Cieek and Back Creek through this gap perhai)S without a tunnel, or, if with one, a very short one. But for a sufficient supply of water for a canal between those streams, we should have to resort to a long feeder from Greenbrier River, and one, too, perhaps, with one or tw > tunnels. Table HI shows the quantity of water afforded by the different streams in this sec- tion of the country at the time I made the examination of it. Bitt we could not depend on anything like those quantities during the greater portion of the summer season. From a mere examination of the table it will be seen that at the time the different streams were measured there was a sup[)ly of water afforded by them more than neces- sary to effect the d sired communication. But having since seen many of these streams whilst employed upon the examination of a route for the Baltimore and Ohio Railroad, (during the dry season,) I am fearful there could not be had at all seasons of the year a sufficient supply of water for a canal communication by this route. But even if there could be found a sufficient supply, it would not then compare with the route from Covington to the mouth of Howard’s Creek, as will be plainly seen from the fol- lowing facts : On this route the length of canal would be greater, by at least 100 miles, than by the Covington route; that is to say, the distance (following the probable direction for the location of a canal) from the Greenbrier Bridge to Huutersville, on Knapp’s Creek, and from thence over the dividing ridge, and down Back Creek, &.c., to Covington would be at least 100 miles, and probably more. 1|[ffie fall of the ground from the western extremity of the summit-level of this route to (the Greenbrier Bridge, near the mouth of Howard’s Creek, is about 1,500 feot. So APPENDIX Y. 801 'iy tbe fall from the eastern extremity of Pocahontas snmmit to the month Creek is equal to 1,970 feet, making 2,552 feet more of lockage on the Poca- to make a comparison between the lockage on this route and that on the Covington rente, wti have the amount of lockage from the western extremity of the summit-level on Covington route to the mouth of Howard’s Creek equal to 224 feet, and the ar^uunt ( f ''ockage from the eastern extremity of said summit-level to the mouth of Dunlap’s C. i k equal to 694 feet, making the amount of lockage for the entire distance from theAioulh of How.ard’s Creek to the mouth of Dunlap’s Creek equal to 918 feet; conseqn^ ofDun®^* hontasM^dc than on the Covington route The Weat additional number of aqueducts which would be required on this route would Flake the cost of its construction very great when compared with the Coving- ton r. u'le. In the distance from the mouth of Howard’s Creek to the mouth of Knapp’s ^pursuing either side of Greenbrier River, it would be found necessary to cross Ho iO of the tributary streams of the Greenbrier; and, owing to the frequent ^oiis in the Greenbrier River itself, it is probable that it would be found neces- b cross it four or five times in some instances to save distance, and in others to e^icountcriug bad ground, &c. The same remarks would apply to Jackson’s ^^o, without a tunnel and with a sufficient supply of water, I am of opinion 'thr. : .ealiontas route could have no advantages over the Covington route. ^Vill liiow conclude by making some general remarks as to the formation of the iltry w'bich we have examined for the purpose of ascertaining the practicability of I'lnection between Greenbrier and Jackson’s Rivers. The rocks in the vicinity of the filing rirlge between these streams are of the transition class, and consist principally /andstone and limestone. Horustone and other rocks arc sometimes met with, but ^'ery sruaU quantities. Of the rocks above mentioned, sandstone is perhaps the most /edominant Compact limestone, (blue,) however, occur in considerable quantities f,nough "it the wlioU' district, and may be Uvsed either as a budding-stone or for the irpose i ffii uishing ordinary lime The most considerable deposit of limestoue which ljav<' '‘Cen u tltis section is a zone or belt which passes trom north to south through [•wislmrg, Union, &c., and which we have traced from its outgoings or appearances ] the surface of the earth for the distance of 20 to 30 miles; its width hardly ever Iceeds 1 . Mj ile. Connected with this deposit of limestone for a number of miles is one found on most of the streams east and west of the dividing ridge, viz, [tiver, on Craig’s Creek, on Dunlap’s Creek, and on Augley’s Creek. In a Me country, from Covington to the Sweet Springs, abounds in limestone. 'Sinking (jreek, on Knapp’s Creek ; on Anthony’s Creek we generally find fn Greenbrier River, above the bridge, we sometimes meet with limestone, formation is of sandstone. here be remarked that the feeder-line from Greenbrier River to the sum- seldom intersect limestone formations, although we frequently findiime- 'icinity of the Greenbrier, on either side, as high up as Huntersville, in suitable for constructions, is found on both sides of Greenbrier River, near of this stone, the piers of the bridge are constructed. id the country, particularly between Howard’s Creek and Crow’s, on Dun- in the direction of the proposed tunnel betw^eeu those streams, and came fiusion that, in the construction of the tunnel for the jiriucipal part of the jgh the ridge, solid sandstone would be met with. jeuever th-s was not the case, whieh would probably be in the commencement ^ nation of the tunnel, we should probably encounter argillaceous slate and 3 with a slaty structure. The i G throughout the whole section of country examined is generally very good, but vai’/'s in quality in a measure with the different kinds of rock mot with, lime- stone sAil being bet ter than sandy soil. CoaUand iron ores are the only mineral productions discovered worthy of much notice. Coal Mas been discovered some few miles from the bridge on Greenbrier River and will probablj^ be found elsewhere, as we have frequently observed formations with which it is almost always associated. We have noticed the appearance of it particularly betweeui the Greenbrier Bridge and Lewisbnrg. We have not, however, seen any s^jeci- meiis of a good quality. This section of eountry is, unquestionably, the repository of numerous beds or de- posits o'liron ore. We are assured of this fact by its appearance in several places. A cousidefiVde lead of it is found on John’s Creek, near New Castle. It also abounds on Dunlap’s Creek and its branches, in the neighborhood of Covington, and will, no doubt, be frequently met with elsewhere. The ore. on Dunlap’s Creek is the red oxide and is considered of a most excellent quality. All of which I have the honor respectfully to submit, &c. John N. Dii.latiukty, Lieutenant First Regiment AriUlerg, on Topograi)lucal Duty. 51 E 802 EEPORT OF TflE CHIEF OF ENGINEERS. REPORTS ON IMPROVEMENT OF THE GREAT KANAWHA RIVER BY LOCKS AND DAMS, BY „ MR. JOHN A. BYERS. \ 1 . Wingfield, Fehrnary ^ 18G?!f. Dear Sir : I have sent to yonr address an estimate of the cost of imp the Kanawha River by locks and dams from Loup Creek to the Ohio River. s I have also sent an estimate for an improvement by locks and dams froin I Creek to Brownstown and from thence down by low dams at the head of the sL i, side- canals along at locks at the foot of the shoals, each of these plans to seciu 5 feet depth of water for navigation. In planning any system of works for the improvement of a water-conrec, ^ ■■;rd should be had to the height and character of its floods, and the materials coitl ig the banks and bed of its stream. Passing by occas onal floods of from 50 to 00 feet as beyond any practical cow. tion, the Kanawha River is generally visited by one to several tides of from DO feet every year. These floods sometimes occur when New River for 150 miles, Greenbrier for 00 n Gauley Ri%'er for about the same distance, are in many parts covered with $kroi\ these rivers having an average fall of 10 feet to the mile. New River, taking it in the warmer climate of North Carolina, is sometimes flooded by rains fal ling while it remains dry and cold west of the Alleghanies. The effect under th ese cir stances is to drive the ice along, accnmnlating as it goes, until it is throwi;. down ^ the Kanawha River in a compact mass, and rolled along the bottom of tbe^ river, i'oi ing dams from shoal to shoal, the destrncrive power of which should not b e overlookt Immense quantities of drift often carried with the flood even trees of the largest siz with the roots attached and dragging on the bed of the river. The banks of the Kanawha River range from 40 to 60 feet above low waier, and p made up of clay, sand, and vegetable mold mixed in every proportion. The river -b is nearly everywhere covered with a close and strong pavement of bowlders tilled between with small stones and gravel. Passing below this rocky pavement, which is from 3 to 5 feet in thick alternating strata of gravel and sand, of pure sand, blue clay, and quick generally far below the levels required for foundations. Solid rock is exposed at Johnsou^s and Red House Shoals at a few i water, extending only across a portion of the river, where it sinks abru; inaccessible for building upon. I found solid rock at Peeled Maple and Arbuckle Shoals along both short’ at a very uniform depth of feet below the surface of low %vater, and Ij rock will be found in other places along the shallow reaches of the river;’ Considering the character of the river as above described, I have estim:; ing entirely on artifleial foundations. The walls of the locks, dams, ai; are to be of masonry laid in hydraulic cement, the materials and workm the best of its kind, with no expenditure merely for the sake of appearai\ prices applied, except for mortared masonry, are prices now paid for simi. progress on Coal River. In the estimate you will notice what may be considered a large sum for o' &-C., but, considering all the difficulties to be encountered, I cannot believe materially in excess. Coffer-dams will be necessary to the proper execution of the w’orke, and be carried 4 to 6 feet above the ordinary low water, so as to render all tne season for founding the locks and dams available by being above the ordinal floods. It must be thought of, that, while these locks and dams are in progress, 1h^ tion of the river will be often obstructed and cccasionally suspended. To eecnomlze these difficulties the work will have to be pressed, so as to shorten the obstrueJion to navigation, by working night and day, all tending to increase the cost of Ih.e .vork sh' uld ' .'.ilding' i)U miner ESTIMATE SHOWING THE COST OF IMPROVING THE KANAWHA RIVER FROM LilFP CREEK TO THE OHIO RIVER, By nioriared-masonry locks and dams. 12 locks, at $68,077 $816, 024 12 dams, at $74,179 r:90, i i8 Channels of approach to the locks 100, 000 Coffer-dams, pumping, &c 60,000 1.867,072