At the annual meeting of the New England Association of Gas Engineers, Mr. Thomas, of Williamsburg, made the following remarks on this subject: "In my early experience with the Williamsburg Gaslight Company, with which I became connected in the year 1854, I found pretty nearly all the street mains that were laid were connected with cement joints. While there is no doubt in my own mind that a joint can be made perfectly tight with cement, I much prefer the lead joint. Another thing to be taken into consideration to keep tight joints is that the mains should be laid a sufficient depth under the surface to protect them from the action of severe frosts. A great many of the mains were not more than 18 inches or 2 feet below the surface of the streets, and at this depth in our climate it is a matter of impossibility to keep joints tight, as the action of the frost in winter will displace the mains and cause the joints to leak. From the bad manner in which our mains were laid, and the cement joints leaking so much, we could not afford to turn gas on during the day. Had we done so we should not have had any to supply the city at night, and we were thus compelled to shut off the gas just as soon as there was any apology for daylight, and keep it shut off as late as possible in the evening.

"With the most careful working in this manner, for a period of nine or twelve months, our losses from leakage amounted to about 52 to 55 per cent of the gas manufactured. A great part of this loss was caused by the cement joints leaking, and also a part due to the fact that the mains were not at sufficient depth under the surface to protect them from the action of the frost. As soon as we possibly could I went over the whole of our mains (there was about 17 miles in all), stripping them, cutting out the cement, and rejointing them with lead. In one season we got the loss from leakage down to 20 per cent, and this with the gas turned on during the 24 hours of the day.

"One great objection to cement joints is the rigidity of them; in cases where pipes have been disturbed by other excavations and settled, I found in all cases that the mains were broken. In a leading main from our old works, with cement joints, the main, a 10-inch one, was broken entirely off and fractured lengthwise besides, by the upheaval of the ground from frost. In some of the same mains that we had rejointed with lead the mains were drawn apart, drawing the lead out, but with very little loss of gas, as the gasket being driven in tight prevented any great leakage. In cases of this kind the lead was easily driven back, and the joint made perfectly tight again. I have never in our city put in any street mains that I have not used lead in the joints, and in laying mains we always make them gas tight with the gasket used.

"At the present time we have over 90 miles of street mains laid, and outside of our loss from street lamps (we get paid for three foot burners and they average about 3¼ foot) our loss from leakage will not exceed 6 per cent. We have suffered severe loss of gas from sewering in our city. In some cases where there are railroad tracks in the streets, the sewers have been run on both sides of the street, alongside and parallel with our pipes; these excavations are much deeper than our mains lie, and the earth is always filled in loosely and left to settle.

"In cases of this kind, whole blocks of mains were dragged down, the pipe broken, and the joints partially pulled apart; at the same time the leakage from the joints was not so great, the gasket preventing the leakage. In laying street mains, what you want particularly to attend to, and especially in the East here, where you have colder weather than we have (we have not seen much winter until we came on here), is to get them down under the surface a sufficient depth to protect them from the frost. With us the least depth is 2 feet 9 inches under the surface of the street, and I am confident, could our mains remain in the ground as we put them down, our loss from leakage by them would be very small indeed. While, as I stated in the beginning, I have no doubt that a cement joint can be made tight, I can see no benefit in using cement for the purpose, as I consider lead far superior in accommodating itself to any upheaval or settling of the earth where the mains are laid down."