As to the question whether a round chimney is not the best, I think there is no doubt of that for the reason that a round chimney offers less frictional resistance than a square one; but as we can make water flow through a square tube just as well as through a round tube, so a square chimney can be used provided you make it a little larger in area than a round chimney.

Mr. B. H. Carpenter:-I should like to ask, in the case Mr. Kent speaks of, whether if that partition had been taken out part way it would have been as successful? I ask this because in one case where we were putting in a boiler there was a chimney of that kind, except that at the base it was separated probably 15 feet and it ran together at the top, and the combined area was hardly large enough for the boiler. But the owner thought that he would run the risk and make the trial and if it would not do would build a chimney outside. We cut the division out for probably five or six feet down from the top and did not have any trouble.

Mr. Kent: I would say that any length that you cut out of that partition would be a help and the farther you cut it out the better. If it had been cut still farther down a little more draft would have been had. The partition simply acts as a damper where it is placed.

Mr. Gormly:-I would like to give the gentleman a little experience I had with a chimney. We put in a steam heating plant in Utah Territory, and I thought I had a direct flow all the way out, but when we undertook to fire up we could not get a draft. After considerable investigation I discovered two flues built side by side, which we thought were separate and distinct flues. We found that there was a communication between these flues about three feet long and probably eight inches wide. The result was that instead of drawing through the grate we found that there was a draft down one of those flues and through this opening at the bottom and into the other flue and out. When we closed off the top of one of those flues we then had an excellent draft, using one chimney and leaving the chimney broken out as it formerly was. We had another case only last week in which we were unable to raise over five pounds of steam in a place where we wanted to run twelve. After considerable investigation we found that at a number of places in the chimney the brick was laid so loosely that it would blow a candle flame right in when there was fire in the heater. I thought that as the discussion has run on to the divison in chimneys I would say it is rather a delicate thing to attempt to run two flues with a partition between them. That has been my experience.

Mr. Cary: In the matter of flues, discussion can be continued

without end almost. The subject has come before the Mechanical Engineers and I don't know that they were much nearer the end when they finished than before they started. As to the matter of a chimney being four inches wide and some 24 to 30 inches the other way, I think the success of that chimney should be attributed to a very small boiler being used. The area was in excess of the area theoretically required by Mr. Kent's formula. This being an outside wall probably the gases as they entered were condensed by being chilled and that, of course, affected the chimney draft. The higher the temperature up to the neighborhood of 550 degrees the better draft you get, but the gases were probably condensed and then, after that condensing took place, there was probably temperature enough to produce all the flow of air that was required. I think there are a good many inexplicable matters relating to chimneys. In Mr. Kent's treatment of this subject he does away with some of the more complicated formulas in taking care of friction, by allowing for it a 2-inch space all around the inside of the chimney; that is, taking the radius of a round chimney, for instance, and drawing a concentric circle in the chimney with a space two inches all around; the space within the inner circle is considered a draft area free from friction. This matter of friction in a chimney is one to be avoided. I know of an instance where the chimney was peculiarly constructed to serve a double purpose. In the center of the chimney a wrought iron flue was built running from the basement, where the boilers were located, to the top. Outside of this it was a square flue and it was tapped into each floor to serve for ventilating purposes to draw the air in from the various heaters all the way up, and it served for that purpose for a number of years. They were obliged to increase the size of their boiler plant and they thought that they could use the inside flue as well as the outside one and get all the area wanted. But when they got the boiler they did not get the results, and they were in a great deal of trouble until they pulled out the interior lining of the flue. Then they got all the draft they wanted. This is a case that comes up when you get bad results by having divisions coming through parallel flues, but by trying to use the two parallel flues individually, on account of the friction, you will not get good results.

Mr. Kent:-In regard to that flue four by 30 inches all the way up, my formula would show that it would not draw at all, because my formula allows two inches all around for friction, and taking two inches all around means no area.

XXII.-TOPIC No. 21.

In a 1-pipe system of steam heating, which works better in practice-a sealed tank return or a vented tank return?

Mr. Barron:--That is the I-pipe system, as I understand it. That represents pressure and exhaust steam combined.

The President:-I should think so.

Mr. Barron:-I think that the sealed tank is the only proper way. You do not want to waste any heat, and there is no difficulty in making a sealed tank work successfully with a pump governor or a pump without a governor.

XXII.--TOPIC NO. 22.

What is the difference between convection and radiation in the number of heat units given off per square foot of surface by steam under five pounds pressure?

Prof. Carpenter:-I had occasion to look that up in connection with my book on Heating and Ventilating Buildings and the only experiments I could find were by the French writer, Péclet. I gave those very fully in my book. You will see that they depend largely on circumstances. That question could not be answered without taking into account a great many things. As his experiments seem to agree very well with our recent ones I think his results are quite reliable. There does not seem to me to be any relation in these experiments unless you assume some condition which is not given here or else determine what your radiating surface will be.

XXII.-TOPIC No. 23.

In what location are the best results obtained from hot air flues -outer walls or inner partitions?

Mr. Wolfe:-We pretty generally have understood up to within a comparatively short time that the location for a warm air flue. should be in the inner wall. My own opinion is that it does not make any great amount of difference where you are bringing the warm air from in warming a building by an indirect system; no matter how you heat the air it will naturally rise to the ceiling. Naturally it will drift to the outside or exposed wall, and the more pertinent question is where shall we take it out to cause the best distribution of the fresh air throughout the room rather than where shall we let it in. We all know that if the register is in the floor in any part of the room that the warm air will go to the ceiling first. Its drift will be-provided there is an outlet-first to the outside surface; from that point the drift will be naturally towards the outlet,

getting its power from the power of the outlet and the additional power of the force of the inlet and the increase by gravity or its extra weight in the fall. The Massachusetts inspectors recommend that the inlet should be adjacent to the outside wall on an inner wall. The objection to a floor register is that it does not add anything to the purity of the air to have it pass over the clothing, shoes, etc., of people who will stand on it. If it is low down it comes in with a velocity which makes it disagreeable to sit near it. Consequently, think that a warm air flue or an inlet flue can be placed anywhere above the seven foot line and insure a pretty good circulation. But it is more important, in my opinion, as to where to take the air out than as to the point to bring it in. Permit me to locate the outlet of proper capacity and power, and you locate the intake, and I can warm and ventilate the room, but with control of inlet only I could not be sure as to results.

Mr. Jellett:-My own experience is that a modern house is a house without doors on its inner walls; that is, there may be sliding doors that are generally open, so that you have to consider the house as a whole. The hall being an open space leading to the upper floors of the building, there is a greater aspiration in the halls than in any other part. The average house, also, is not provided with outlets. There are very few residences to-day that have a complete system of outgoing flues. These being the known conditions under which we are asked to work, I would prefer to get my flues in the outer wall for two reasons. The first one is that as houses are constructed we can get a cheaper and more effective flue. The second one is that the form of hot air register used to-day will shut off a larger proportion of the area of the flue than we can get into an inner partition, unless we use an extended register, and they are considered very unsightly and are generally objected to on that ground. The average partition is a four or five inch partition. The average register box is 2 to 21 inches in depth. If you allow for plaster on the back of your flues you have less than two inches in depth to your flue and therefore your flues are cut down to a very small area. The studding is placed twelve to fifteen inches. apart, so that you are limited to four inches in depth and an extreme of sixteen inches in width. It has been our practice in doing work of that kind to place flues in outer walls, to make a hollow backed flue and fill the hollow back of the flue with mineral wool or something of the kind. I have in mind a case of three large hospital buildings where I made the plans for heating and ventilation. It was absolutely necessary to lead through the outer wall. The hollow backed flues, such as I have described, filled the space. We

got very satisfactory results with flues erected in this manner, but we did not find satisfactory results with flues on the inner wall from the fact that we could not get sufficient area to make the flue effective.

Mr. Connolly:-I do not agree with Messrs. Wolfe and Jellett, having had an extension to a city house with three bath rooms, second, third, and fourth floors, heated by indirect hot water. Upon the advent of cold weather the house was thoroughly heated, with the exception of the third and fourth floor bath rooms, which were down to a temperature of 50 degrees. The separate pipes for those two rooms were placed on the outside wall, with a double back packed with mineral wool and four inches to the weather. The second floor bath room with a pipe in the inside wall was at a temperature of 70 degrees; each pipe was supplied from a separate compartment of a stack of indirect radiators in which circulation was thorough, and special provision for ventilation in each bath room was made. To improve the heating of those two rooms an 18-inch fan was erected to the mouth of the cold air inlet, disconnecting the power wire from an electric house pump not in use, and connecting it with this fan; after a fair trial of forcing air through the stack the third and fourth story bath rooms were still uncomfortable. As a final remedy an auxiliary stack of radiators was placed at the base of those flues, using the heated air from the original stack as a.cold air box for the auxiliary stack, thereby superheating the air. After this change the fan being started, the two bath rooms were heated; so from that experience I would not advocate hot air flues in outside walls.

Mr. Jellett:--In answer to the last speaker I would say that I never put a flue in with only four inches of outside wall. It is against the law in our section of the country to come within that limit. We must have nine inches of back always on outer walls. Four inches of brickwork is little or no protection to the flue. The outer walls have a certain thickness in proportion to the height, so that usually we find that we never have less than nine inches back of the flue.

Mr. Connolly:-All I would say in answer is that in ordinary buildings the walls on third and fourth floors are not over twelve or eight inches and naturally there would be only four inches to the weather. They may, of course, build walls thicker, and use more brick in Philadelphia than in New York.

Mr. Wolfe:-Mr. Chairman, I only want to explain my position in the matter. I have had to do with but one dwelling house in ten years. When I speak of my experience, it has been entirely