application, as circumstances require, which is usually one part of the concentrated chemical to 18 or 20 parts of water, with the arsenic-caustic soda combination referred to above. The application is made on any desired scale from a hand car, single flat car, or entire train, all with proper equipment. The fully equipped train is the only method with which to work economically on a large scale. The train is made up of 6 or 8 cars. The front or pilot car of the train contains the sprinkling apparatus, composed mainly of pipes, perforated with small holes or nipples for the emission and diffusion of the poisoning liquid, valves, meters, etc., all under the complete control of the operator. In the same car are pumps and air compressor actuated preferably by motors of Marine type, to furnish pressure to the liquid, in applying it to spraying the weeds, and to furnish compressed air for mixing the diluting water with the concentrated chemical. The liquids are thoroughly mixed by means of compressed air before application. In the after part of this car may be living quarters for the crew. The main part of the train is made up of tank cars-all connected together with lines of pipes, and fully supplied with valves, meters, all necessary appliances. One of these tank cars is loaded with the concentrated solution which is drawn out as needed. The other tank cars carry a supply of the chemical diluted to the strength for spraying on the weeds. The last tank car contains Repellant, a liquid to be sprinkled over the weeds after the poison has been applied at road crossings, station grounds, etc., in places where cattle are liable to browse on the vegetable growth and be poisoned by the compound. This Repellant is bitter or contains something disagreeable to grazing stock, this however appears to be of doubtful value. Last of all comes the locomotive pushing the trains. Before starting to sprinkle-the tank cars carrying the diluted mixture are filled. The concentrated chemical is drawn from its storage car and mixed 14 to 20 parts more or less of water as practised by the company with largest equipment and doing most treatment on the roads, while another company dilutes with equal volumes of concentrate and water. Thus the amount of dilution must be determined in each case from the strength of the solution and the vigor of growth and thickness of the stand and kind or species of weeds to be killed. The volumes are measured by meters and gages. The tank cars ought to contain enough liquid to reach to the next water station, 20 miles or more. The solution is applied at a speed of about 20 miles per hour, and by following a freight or local train as a second section, there is no interference with traffic. When the next water station is reached, a stop of an hour is made-although half an hour is usually long enough to refill the empty tank cars with another supply. The rate of application in volume per unit of area varies according to condition as much or even more than the strength of solution. A very convenient unit of measurement is gallons of solution per foot of width per mile of roadbed, which appears to be ordinarily within the limits of 5 and 15, or a minimum application of 5 gallons for width of 8 feet will take 40 gallons per mile, while a maximum application of 15 gallons per foot for 14 feet width will take 210 gallons. The rate of spraying per unit of surface is regulated either by controlling the amount of liquid allowed to escape from jets, which can be made under the control of the operator, or by varying the speed of the train or by both means. The rate of emission should be directly proportional to the weed growth and must be controlled by the judgment of the operator. It is claimed by some roads using the chemical weed killer that the weeds should be allowed to attain their full growth before application, so that the sap in descending the stocks will carry the poison to the roots. Others claim that the chemicals should be applied to the growing weeds in order to eliminate them from the roadbed during the summer season. It should be applied before cutting the weeds, as cutting causes the weeds to seal over the cut, preventing the poison penetrating to the roots. It should not be applied immediately after a rain as the cells of the weeds will be full of water and prevent much poison entering. It is claimed in some reports that the chemicals should not be applied immediately before rain, as rain washes the solution away so it will not enter the plant. It is claimed by others that rain after application is beneficial, because the solution being dried, is redissolved by the rain and is then absorbed by the plant. It is claimed by the advocates of chemical weed killing that successive application in lessening quantities per year for three years will completely extirpate the weeds, and the only ones afterwards growing will come from seed blown or carried in on the right-of-way. It is also held that the ground may be completely sterilized by the chemicals so it will grow weeds, but no instance of it actually being done has been given. Only one case has been reported so far of complete killing of weeds after three years' application. Some reports indicate that the most effective method is to make the first application so strong that all the weeds are killed-that successive weak applications tend to make the weeds immune. Some grasses and weeds, such as Johnson, Burmuda, Salt, Blue Stem, Wire, Crab Grasses, Horsetail, Ragweed and Bull Thistle, do not appear to be killed by a single application, and their complete destruction is still a problem. One road reports that after four years' application it was decided to discontinue entirely the use of the weed killer and resort to the weed burner, claiming that the "weed killer" did not destroy all the weeds and those that did succumb were dried and left to foul the ballast and otherwise giving an unsightly appearance to the roadway, and also that the cost was far in excess and also the results less satisfactory with the weed killer than with the weed burner. Attention is directed to the application of common salt to weeds on 454 miles of the Erie Railroad, as reported by Mr. I. H. Schram, Regional Engineer. The salt was applied from the car by shoveling on chutes which made the distribution in work train service at the rate of eight tenths car loads per mile. The cost was for labor and train $13.40 per mile, for foreign freight $18.75, total $32.20. The cost of hand weeding in that district has been $73.33-which gives cost of salt application 44% of cost of hand weeding. The salt was good mined salt, but rejection from crusher account being too coarse for market, and was received without cost. Approximately 40 tons per mile was used. The application was very effective and killed all the weeds. In this part of New York, in the vicinity of Rochester, salt has been used for a number of years for weed killing. Care, however, must be exercised in its application to prevent interference with automatic signaling. CONCLUSIONS Your Committee does not believe it has as yet enough data to report definite conclusions. It has many favorable reports of successful application and some of unsuccessful, or at least unsatisfactory ones. The reports nearly all cover only one year's work-the results of at least three successive applications ought to be definitely known before framing conclusions. The railroads that have been using it are generally continuing it on a yearly enlarged program, although a few have discontinued. There are cases where the contractor has been asked to do the work over again account unsatisfactory results in first application. How much this is due to unfavorable circumstances and how much to the system itself has not yet been ascertained. The reports so far received indicate that a very large percentage of the soft annuals or tender weeds are killed by the first application-running in many cases up to 85 or 95 per cent, but a smaller proportion of perennial and hardier annuals only are exterminated; about 50 per cent more or less so far as reported. The Committee can only report considerable work done and little progress for the past year, and would recommend a continuation of the study-as several roads are now gathering information which will be of value on the subject for future report. Appendix H (8) DESIGN AND USE OF REINFORCED CONCRETE W. C. CURD, Chairman; CULVERT PIPE H. B. ROBINSON, Sub-Committee. Your Committee has consulted with the Masonry Committee and are advised the above subject was assigned them in 1917. Progress reports and tentative specifications have been submitted by that committee and will be found on pages 748 to 762, Vol. 19, and pages 697 to 701, Vol. 20 of the Proceedings of this Association. In 1919 the Masonry Committee reported that "while progress has been made they feel that further investigation as to the distribution and transmission of pressures from engine loads and earth fills is necessary before a conclusion can be reached." Other Societies, having the same subject also under consideration, appointed a Special Committee in 1920, composed of two members from each of the following societies: American Society of Civil Engineers. American Concrete Institute. American Highway Association. American Concrete Pipe Association. American Railway Engineering Association. This Special Committee met the same obstacle-the lack of information concerning earth pressures and pipe loading-and are now awaiting results of experiments at Iowa Agricultural College before reporting, which experiments will probably require another year to complete. Your Committee has confined its investigation to the application of reinforced concrete pipe and finds it of general use among railways. The majority of the roads prefer to purchase under approved specifications from reputable manufacturers rather than to attempt its manufacture. It is reported that the principal advantage is its cost, particularly under war conditions, when it could be obtained at a figure greatly under that quoted for cast iron. The greatest objection to its use is the excessive weight of the individual piece, chiefly in the large diameters, which often necessitate work train and derrick for placing, thus increasing the cost. Preference is expressed for the circular section rather than for oval section, as in many cases the oval section has been laid with flat side down, which resulted in failure. The circular section, properly designed and manufactured, eliminates this difficulty. Some favor pipe culverts only at unimportant openings and limit the maximum diameter to 36 in., as above that size a concrete box, affording a greater waterway, may be constructed at practically the same or less cost. Your Committee concludes that where cost and other essential items are equal, there is no particular advantage in using reinforced concrete pipe in preference to other approved forms of culverts. In view of the joint investigation of the subject now under way by a Special Committee of other societies, your Committee recommends the Masonry Committee be permitted to conclude the subject for this Association at such time as may be proper, and that your Committee be relieved of the subject. Appendix I (9) THE EXCESSIVE COST OF MAINTENANCE DURING THE EARLY PERIOD OF OPERATION C. C. CUNNINGHAM, Chairman; W. M. JAEKLE, R. C. GOWDY, W. H. WOODBURY, Sub-Committee. There has been considerable amount of thought spent on the part of the various engineers of the various railroads of the United States relative to the question of the increased cost of maintenance of roadway and track during the early periods of operation of newly-constructed lines. This question involves a good many factors and must be carefully studied as it is applied to railroads built in different localities. In order to develop information as to whether the cost of maintenance is greater during the early periods than it is during the later |