too technical for any proper treatment here, and would require a careful examination. I think it is, therefore, almost impossible to add anything of interest.

In a city of the general size of Washington, D. C., the use of the alternating current for railway service did not at all seem desirable, until the question came up of suburban roads, which roads in one case were sixteen miles, in another case twelve miles, and in another case eleven miles from the center of the city, and when those cases arose it did seem to be the fact that the alternating current should be used for transmission on those outside lines; in the meantime keeping all the inside business within the direct service to a distance of about seven or eight miles. I think that this fact may throw some light on the subject.

Mr. Hopkins-I might state, in reply to the remarks just made, that the distances from the main station to the various substations are given in the paper; the shortest distance being two miles to the Milo station, where it is intended to install a substation to replace the present steam plant. We now have two steam power plants which are operated most of the time; our idea being to locate the rotary at this point, principally to enable us to discontinue the steam plant.

President Holmes-I will appoint the following named gentlemen as the Committee on Nominations to propose the name of officers for the ensuing year, and also to recommend a place for holding the next meeting:

Mr. Charles S. Sergeant, Boston, Chairman.

Mr. Albion E. Lang, Toledo.

Mr. Frank L. Fuller, Wilkes-Barre, Pa.

Mr. Walter P. Read, Salt Lake City.

Mr. W. W. Wheatly, New York.

It is now in order for any gentlemen to extend to this Association an invitation to visit their city the coming year. It has been customary, at our former conventions, to have those who care to invite us to be their guests for the coming year, to do so and have it go before the Committee on Nominations.

If there are any such here this afternoon, we shall be glad to hear from them.

To-morrow we set apart to inspect the exhibits. There will be no meeting in this hall until Friday morning; and in view of the fact that we have a large amount of unfinished business, I think it would be well to begin our Friday morning meeting promptly at 10 o'clock, and if there is no further business we will adjourn until that time..

FRIDAY-MORNING SESSION.

President Holmes called the meeting to order at 10:15

a. m.

President Holmes-The first business this morning will be the consideration of the paper on "The Best Manner and Mode of Conducting the Return Circuit to the Power House," by Mr. E. G. Connette, Vice-President and General Manager of the Syracuse Rapid Transit Railway Company, Syracuse, New York.

Mr. Connette read the paper:

THE BEST MANNER AND MODE OF CONDUCTING THE RETURN CIRCUIT TO THE POWER HOUSE.

The American Street Railway Association

Gentlemen: The writer is somewhat at a loss for new material to offer on this subject, as it has been discussed for years and improvements introduced until the present practice of using the rails with bonded joints for conducting the return circuit has been practically perfected, so far as the use of the rails of the track can be utilized for that purpose; especially when the highest state of the art of bonding the joints is particularly observed and the work carefully done. A large portion of the troubles which exist under the present practice is on account of inefficient, careless work, and the use of faulty material, but even when the greatest care is observed and the best material is used in bonding the joints the deflection of the joints in the course of time from various causes will impair the efficiency of the ground return.

A few years ago when electricity as a motive force had been perfected to such a degree as induced the street surface railroads

to transform their motive power from horse to electric, there were two electric companies in the field offering apparatus for street car propulsion, viz., the Sprague Electric Company and the ThomsonHouston Electric Company. In 1890 the writer assumed the management of a street railway company, a part of which was being equipped by the Sprague Electric Company and the other by the Thomson-Houston Electric Company. The Sprague Electric Company used the rails of the track exclusively for conveying the return circuit, employing a No. 6 galvanized iron bond riveted to the rails around each joint. The Thomson-Houston Company used an auxiliary copper wire, the same size as the trolley wire, laid in the center of the tracks on the tops of the cross-ties, connecting the same with a wire of a smaller size with a rivet to the center of each rail, and I believe it was their idea to use the same amount of copper wire for the ground return along the tracks, connected in the manner as above described, as was used overhead. In a few months it was discovered that the voltage on the lines equipped by the Sprague Company was very low, especially toward the end of each line, and upon investigation it was found that the galvanized iron bonds had almost entirely disappeared, while on the lines equipped by the Thomson-Houston Company the loss in potential was about the same as when the work was first installed. The rail joints of the Sprague lines were rebonded with copper bonds riveted to each rail, but in the course of time it was further discovered that these bonds were too small in size, and that the loss of potential was gradually increasing on account of the bonds deteriorating and being broken off, and it was necessary to go over the lines again and rebond them. On account of the large initial expense of an auxiliary ground wire and of the improvements and progress in the method of bonding the joints, and the use of heavier rails, together with the improved method of supporting the joints, the use of the auxiliary wire in the ground for the return circuit was abandoned.

With the great variety of devices and improvement in the art of bonding the rail joints at the present time, there is yet more or less trouble experienced, and the tendency is toward a jointless metallic return by using electrically welded or cast iron joints, which practically makes an unbroken metallic return so far as the tracks are concerned; but, unfortunately, a large number of street railroads are not financially able to relay their tracks with heavy rails and use the improved method of joint support and connections, and the problem with them is yet unsolved as to how to bond the joints of the rails so as to get a perfect connection and one that will stand and overcome all resistance or loss. It must be remembered that in a mile of track there will be from one hundred and seventy-six to

three hundred and fifty-two joints, and while the rails of the track have more than ample capacity for conveying the current, the joints inust be so connected as to give the current a path across with but little or no resistance. When the joints are properly bonded the track for a period of time answers well for a return circuit, but after a while the connections at the joints become corroded, the bolts are worn, the joints deflect, and the bonds here and there are either broken or worn in two, or corroded and loose at the connection, and the result is a very imperfect and inefficient ground return. In some instances there is an insufficient amount of wire used for the return circuit from the tracks to the power house, and in some cases when there is enough wire for this purpose the connections to the rails are inadequate and create a resistance that neutralizes the conveying capacity of the copper intended to convey the return circuit to the power house. The wires of the return circuit should be connected to the rails with a connection of equal capacity to the conductor, and should be connected in three or four places, so that if one should become broken or disconnected there would still be ample carrying capacity in the other connections.

In the early days of telephone exchanges the ground was used as a return circuit, and we are all familiar with the inefficient and poor telephone service while this state of things continued. The service was disturbed by the trolley currents and other influences until finally the telephone exchanges were forced to put up a metallic return in order to save trouble, annoyance and disturbances to their service, and I believe after a trial it was discovered that the expense of putting up the metallic return was more than compensated for by the improvement in the efficiency of their service and was money well invested.

It is, therefore, the opinion of the writer that "the best manner and mode of conducting the return circuit to the power house" is by using practically a jointless and independent metallic return, connected to each rail, and the connection made with ample surface contact and absolutely water tight. On account of the chemical action which is likely to take place between the copper rivet and the steel rail, the rivet should be covered with a thick coating of lead, solder or tin. The best connection which can be used, if the facilities are available, is to electrically weld the connection from the auxiliary ground wire to the center of each rail, and, I dare say, that even with the extra initial cost of an independent ground wire it will result economically in the long run. If an independent metallic return circuit was the custom now, there would be no necessity for a paper to be written on this subject; and there would be no more need for discussing the subject of rail bonding, or of electrolysis, or of the

best method for discovering bad connections in the return circuit. Nor would there be the necessity for the expenditure of a large amount of money for expert investigation and advice to cure the bad results of the present imperfect system of conducting the return circuit to the power house; and the amount of money saved by loss of power, cost of rail bonds and labor of renewals, tests and investigations, as well as the trouble and expense incident to electrolytic action, would be more than ample to pay a large interest upon the cost of installing an independent ground return.

Cast iron joints and electrically-welded plates on the joints make the rails of the track practically a jointless metallic circuit as a conveyance for the electric current, but the very large number of joints in the rails of the track make the certainty of an unbroken and continuous metallic circuit a doubtful proposition, as the wear and tear of heavy traffic and the expansion and contraction of the metal will, in the course of time, produce results which may very materially impair the efficiency of the joints as a perfect conductor, and it appears to the writer that it is just as essential to have a conductor of as perfect design and as efficient in carrying capacity, to convey the current back to the power house, as is used to convey the current from the power house to the car, and if this were the case, there would be no more trouble with the return circuit. I do not advocate a double trolley, because of the difficulties which are well known to street railway people, but an independent metallic return used in connection with the single trolley system is not as expensive to install as the double trolley system, and, practically, obviates all the troubles incident to the ground return where the rails with bonded joints are used for the return circuit. Respectfully submitted,

E. G. CONNETTE.

President Holmes Mr. Roberts, will you discuss the paper?

Mr. Roberts-Unfortunately, I did not get in at the starting of the paper and have not received a copy of it.

President Holmes-I do not think there are any copies printed; consequently, Mr. Connette had to read the paper. We will be very glad to hear anything you have to say upon the subject.

Mr. E. P. Roberts, Cleveland-The matter of bonding is a serious one to all of us. I have been connected largely with interurban roads of considerable length, and I very much question whether the plan to have return wire conductors through