1 Ton Tare=1.44 Rating Tons.

Total Train = 1931 Rating Tons.

Behind Tender 1744 Rating Tons.

FIG. 5.

Grades out of a divisional point should be reduced to a somewhat lower grade than at other points, as the first few miles a train is run after standing for some time show a very much higher resistance than the normal, especially in cold weather. The writer knows of one case, in extremely cold weather and high wind, but with pretty clean rails, when it was impossible to start a train out of a level, crooked yard with the locomotive unassisted, but the same locomotive, an hour afterward, handled this train on a 0.4% grade as usual.

For ordinary stops, at least 0.1% compensation is advisable, although a compound locomotive can generally start anywhere that it can be run through without losing speed, but at greater risk of broken draft rigging and loss of time, to avoid which generally justifies a slacking of the ruling grade.

On favorable divisions, the rating can be raised 50% for about 10% of the original construction cost by the proper introduction of momentum grades, and, incidentally, the rise and fall and curvature be improved. These improvements may often postpone indefinitely the necessity for double track, and are in line with, as well as being a preliminary to, double-tracking. Because of fewer stops, momentum grades generally tax locomotive powers less on double-track lines.

AMERICAN SOCIETY OF CIVIL ENGINEERS.

INSTITUTED 1852.

PAPERS AND DISCUSSIONS.

This Society is not responsible, as a body, for the facts and opinions advanced in any of its publications.

IMPROVEMENT OF THE BLACK WARRIOR,
WARRIOR AND TOMBIGBEE RIVERS,
IN ALABAMA.

Discussion.*

By Messrs. WILLIAM M. HALL, B. F. THOMAS, D. M. ANDREWS and
R. C. MCCALLA.

WILLIAM M. HALL, M. Am. Soc. C. E. (by letter)—The writer has Mr. Hall. read this paper with much interest, and, as the literature on the locating of locks and dams is very scant, it is hoped that the Society will soon have more papers on the same subject.

The writer will not discuss the paper; but, as he has assisted in the location of several locks and dams; and, as he has been engaged during the past year on the location of two Ohio River locks and movable dams, some of the guides which he is using in the latter work may be of interest to others. They are briefly as follows:

1. Select a location where sufficient width can be had for a lock 110 ft. wide, a navigation pass of 600 ft., and a weir of at least 250 ft. 2. Select a location which, with small expense, will admit of the passage of tows around the coffer-dams of the various parts during construction.

3. Select a location where the sill of the navigation pass can be placed above the bed of the river, and, as low as the controlling contour above and below, without danger of the movable parts, when down, becoming buried in sand.

* Continued from October, 1902, Proceedings. See April, 1902, Proceedings. for paper on this subject by R. C. McCalla, M. Am. Soc. C. E.

Mr. Hall.

Mr. Thomas.

4. Prefer a location where the channel above and below will be straight for 2 000 ft. or more; and avoid locations where it will be necessary to "flank" fleets of barges in passing through the navigation pass.

5. By use of core-drills for testing the depth to and quality of ledge rock, and by other careful investigation, select the location, filling the above conditions, which has the best natural foundation within the limits of the stretch of river on which it is possible to place the lock and dam.

6. Prefer a location at the foot of a shoal with a natural pool below and near by.

7. Prefer a location below and near navigable tributaries; thereby improving the tributaries as well as the main stream.

8. Seek locations with good, firm and high banks, and with suitable land near by for power-house, store-house and the tenders' quarters.

9. Prefer a location where the lock can be placed on the opposite shore from drift. (On the upper Ohio, the prevailing winds being from the west, unless the conformation of the river or local tributaries prevent, the wind usually carries the drift near the easterly shore.)

10. If necessary to select a location in a bend, prefer to place the lock next the concave shore, and as near to the head of the bend as other conditions will allow.

11. Avoid cutting the harbors of important towns or cities in twain by a lock and dam.

The writer wishes to acknowledge the receipt of much information on the foregoing subjects from the reading of Department Reports by W. H. Bixby, Major, U. S. A., M. Am. Soc. C. E., and B. F. Thomas, M. Am. Soc. C. E., and by the discussion of these subjects with them and with W. E. Craighill, Captain, U. S. A.

B. F. THOMAS, M. Am. Soc. C. E. (by letter).-All engineers engaged in designing and constructing works of river improvement have felt the need of papers of this character, giving in detail the various parts of the design and methods of construction. Except in a general way, and in isolated pamphlets, there is no work, except De Lagrene's "Cours de Navigation Intérieure," which treats of this important branch of engineering, and this treatise, published some thirty years ago, is now out of print. Then there are numerous Government reports which contain valuable information, but a search for it involves an expenditure of much time and labor; so much, in fact, that a busy engineer cannot undertake the task. This paper, coming at a time when many engineers are engaged in the study of projects for improving rivers by canalization, is a valuable addition to literature of this class; and, owing to the care with which it has been prepared, it will

prove of much benefit to such engineers, and of interest to those not Mr. Thomas. thus engaged.

It is not the purpose of the writer to review the paper in detail, but rather to discuss the subject generally from the standpoint of one who has been giving it pretty close attention for a number of years. Location. The practice of making trial borings for foundations is a most excellent one, and no work of importance should ever be commenced without first making a careful examination of the river-bed material. This has been too often neglected, sometimes with disastrous results. Not only is this necessary in order to determine the character of the structure to be built, but it is also advisable from the standpoint of economy. Investigation, with proper appliances, of all the possible sites is the only method by which the cheapest as well as the most suitable location can be chosen. The borings, where rock exists, should extend into the ledge from 4 to 5 ft., in order to be fairly certain that a boulder or loose rock has not been encountered, and they should be made at intervals of not more than 50 ft. under the important parts of the structures. The water-jet method used by the author is very satisfactory, but, where the number of holes is limited, or where there is an insufficient depth of water to float a suitable plant of this character, the writer has found the method of driving the casing with hand pile-drivers quite economical, and sufficiently rapid for ordinary purposes. A light pile-driver, which can be carried anywhere by three or four men, is set up on the sand bars, or on a lightdraft pushboat, and fitted with a banded wooden hammer of such weight as can be lifted readily. This is operated by four men pulling on a rope attached to it and passing over a sheave in the head of the driver. Two of the men stand on the second floor of the driver and two on the bottom. As the pipe is driven the material is pumped out with a sand pump, the driving being stopped during the operation. When bed-rock is reached, a churn drill is dropped into the casing, and the hole is continued with it.

In seeking his locations in "curved instead of straight reaches" the author has departed from the usual practice, but, when the curve is of very long radius, the idea is to be commended; because in, such locations, with the lock on the convex shore, protection from drift is afforded, which cannot always be counted on in straight reaches. It is not good practice, however, to place a lock having a fixed dam in an abrupt bend, but when the conditions require it (as they sometimes do, because it seems that the best foundations are usually on the concave sides of bends), the question at once arises as to whether to place the lock on the convex shore (the "point "), as the author has done, or on the concave bank (the "sag or bend "), as has been the most general custom. In a river carrying considerable drift, the "point" side is indicated because of the protection it affords, but such location renders