kept 8 to 10 days longer than the others in the dry air of the laboratory, which seems to have had a disastrous effect on them. But in spite of these slight drawbacks, the annexed table shows that up to 24 per cent., the percentage of water has not a very great effect on the strength. This is an important point, for below 20 per cent. the mortar obtained is rather dry and very difficult to handle.

But beyond this limit of 24 per cent. a greater proportion of water seems to weaken the concrete considerably.

This limit is very sharply defined in the adjoining table, where an additional 2 per cent. of water, from 24 to 26 per cent., weakens the concrete by almost one-half for the one-week tests. It is, however, interesting to notice that strength is almost completely recovered with time, the four-week tests showing the weakening limit to be between 26 and 28 per cent., and the two-months' between 28 and 30 per cent. So that if immediate strength be not required of the concrete structure, 28 per cent. of water will not affect the ultimate resistance if allowed to stand two months.

In the parallel sand and cement tests the weak line is not so sharply defined, but yet it is sufficiently so to show that the same statement applies. The tests in this case show a marked weakening between 14 and 16 per cent. of water for the one-week, which strength is ultimately recovered, as is shown by the four-weeks' and two-months' tests.

The low limit of 14 per cent., as compared with 24 for the concrete, is probably due to the fact that the stones of the concrete, on account of their porosity, absorb a part of the water.

The table shows that the greatest density is obtained with 16 and 18 per cent. The weights of the cubes beyond this decrease up to 24 and 26 per cent., where they are again nearly equal in density to the 16 and 18 per cent. of water. Therefore this 24 and 26 per cent. seems to be the point where the best practical results are obtained, because 16 and 18 per cent. make up too dry a concrete to allow of easy handling.

Another point incidentally comes up. Attention has been drawn to the poor results obtained by the same tests and reason of long exposure to dry air given. This shows up a very important point, namely, the necessity of covering up carefully all concrete and cement works exposed for any length of time to dry air and sun. The bad effect of these agents is plainly demonstrated, and it is doubtful whether much strength would ultimately have been recovered.

It is also interesting to notice the results obtained by the concretes made of 1 part of cement, 2 of sand and 5 of stones, and cement, 2 sand and 6 of stones. The specimens of these compositions gave results equal to concretes, 1, 2, 4, showing that for strength they are as good as the ones containing a less proportion of stones, while being much more economical.

These experiments are as yet very incomplete. But it is hoped that the researches in this subject will be continued, and that valuable information for the engineer in practice derived from them.

CONCRETE TESTS-COMPRESSION.

Proportions by weight: I part cement, 2 sand, 4 stone.

Proportion by weight: 1 cement, 2 sand, 5 stone.

* Line of weakness due to excess of water.

McGill University, April, 1896.

THE INCANDESCENT FIRE MANTEL.*

By T. E. PYE, of London.

The incandescent fire mantel is an appliance designed to facilitate the combustion of non-bituminous fuel in the ordinary domestic fire-grate, and thereby to secure a fire of pleasing appearance and thorough efficiency, which shall at the same time emit neither smoke, smell nor any objectionable fumes whatever. Wood of all kinds, bituminous coal of the best or of the poorest descriptions, anthracite coal or gasworks coke, will all burn with facility and intensity in any ordinary grate fitted with the fire mantel, whilst the addition of a limited percentage of so-called noncombustible material, such as spent oil shale, cannel coke, etc. (say from 15 per cent. to 20 per cent.), will not materially deaden or even affect the fire.

Of all the fuels just mentioned, there is no doubt that, in view of the claims made on behalf of the new appliance-viz., a smokeless fire which is maintained at about half the expense or an ordinary coal fire, gasworks coke is the best suited for use under the system, and a varied experience of this residual, as employed in conjunction with the mantel, has clearly demonstrated its utility for all ordinary purposes of domestic heating.

It is not claimed in connection with the appliance that in the domestic grate we find an actually new field for gasworks coke: and further, it is admitted that in some few instances this fuel has been successfully employed by gas engineers and other technical men in domestic warming without an admixture of coal or other artificial aids to combustion. It will, however, on the other hand, be readily admitted that if the said field be an old one it certainly has, as yet, been but little worked upon, and that for all practical purposes gasworks coke used alone is, as a popular domestic fuel, a conspicuous and certain failure.

That if burned under conditions suited to its proper and complete combustion gasworks coke would prove an ideal fuel for such purposes has long been known by the most eminent authorities. Professor Ramsay recently pronounced most strongly in favor of the general use of coke. Speaking upon the smoke nuisance, he said: "There is one material-viz., coke, which gives

*A paper read before the Midland Society of Gas Engineers.

out, weight for weight, a greater amount of heat than coal, which emits no smoke, and which produces what all of us are so unwilling to relinquish-a bright glow in the grate. In Paris it is the universal fuel. Could we induce our dense population to adopt coke fires our smoke nuisance would be greatly abated. Paris is a larger city than Glasgow, and yet it is clean and bright, while Glasgow is dark and dirty. The difference is almost entirely due to the fuel burned."

With this introduction I have much pleasure in bringing under your notice the incandescent fire mantel, believing, as I do, that providing what has for so long been required for the proper combustion of the coke, it is destined to do yeoman service for many a perplexed gas manager in assisting him in the profitable disposal of what for so long has been a positive drug on the market.

It is quite unnecessary to take up time at a meeting of this nature in arguing the importance of the considerations arising from what is practically the development of a new industry, and which, if one-half of the anticipations of the promoters of the system be realized, will almost revolutionize the present condition of coke sales. You will also readily anticipate most of what I may say upon the chemical actions attendant upon the use of the appliance, and I propose, therefore, to ask attention for the present more particularly to such practical issues as have already been decided, and to present data of actual experiences.

The fire mantel is, as you see, a simple contrivance, consisting of a frame carrying two rows of air tubes. The frame is placed inside, and close up to, the front bar of the fire grate. When the fire is in operation these tubes convey into the interior, at two different levels, a supplementary supply of atmospheric oxygen. The carbonic oxide produced in the fire is thus met, as generated, by what may be termed the secondary air, the heat attending its conversion to the dioxide being utilized in raising to incandescence all the fuel in the immediate vicinity of the monoxide flame. By this means a large amount of heat is obtained from the incandescent material for useful radiation into the apartment which under ordinary circumstances is in great part or wholly lost, as the wellknown carbonic oxide flame, usually noted in considerable amount at the top of an ordinary coke fire, although an intensely hot one, gives out no sensible heat by radiation, and its useful heating effect is actually nil. The place and action of the tubes should not be misunderstood. It is not claimed that the air currents toward the fire are concentrated on the tubes, nor yet that the tubes themselves directly increase the air supply. They do so, however, indirectly, as will be shown presently. What the tubes actually do is to convey a portion of the air, while still containing its complement of pure oxygen, right into the heart of the fire, whereas the bulk of what oxygen enters the fire through the front, in the

ordinary way, is quickly converted into monoxide and dioxide of carbon. Herein lies the secret of the brilliant combustion beyond the tubes, at a point which without them, and especially in a small fire, presents an area of most incomplete combustion and halfblack fuel. The intense heat thus set up and diffused in every direction maintains a sharp, upward current of air, which serves a twofold purpose. The heated products of combustion, supplemented by a large volume of rarefied air, and minus the usual complements of smoke and dust and other clogging and weighty matter, possess a buoyancy which, so far as present experience goes, has invariably succeeded in overcoming the worst down-draughts in the ordinary chimney, whilst, as already explained, the CO produced in the fire is already largely or wholly converted into CO2; and the only remaining fumes to be noticed, due to the sulphur in the coke, are thoroughly aerated in the fire and wholly carried up the chimney, and not the faintest trace can be detected in the apartment. Another very important effect of the rarefied air currents is the suction induced at every point of the fire, drawing fresh supplies of air, and maintaining in the apartment a free and healthy circulation, which insures a constant renewal of the atmosphere and the removal of deleterious gases.

With reference to the cost of the fire, my own experience, at home in the sitting room and in the Company's show rooms, shows a consumption of coke costing at present about 3 cents per day, which will not exceed 4 cents in the winter time. Coke is sold in London by measure, but I believe present prices range from $3.87 to $4 per ton, inclusive. From this data you will readily calculate the relative cost of the fire in any locality. You will probably have already seen the figures given in this connection by Mr. Geo. Lane, of Aylesbury, which are as follows: "For 12 hours' consumption of coke it cost, at 8 cents per bushel, 4 cents to heat a room 15 feet by 18 feet, maintaining a temperature of 65° F."

The manipulation of the appliance is as follows: Having filled in the bottom row of tubes the mantel is placed in position just inside the fire bars, and the fire laid in the usual way with paper, wood and coal. The top row of tubes is then inserted and the fire is lighted. When the coals are well alight coke is added, and is then used alone for the rest of the day. Where upright Bunsen gas burners are available they will provide a ready means of starting the fire without the use of coal.

In conclusion, the claims here put forward of a satisfactory solution of the long-standing difficulty in the way of a universal adoption of coke for domestic heating-large and important as these claims may be and are-are daily being supported by highly respected members of your own profession, whose names are a guarantee of their disinterestedness-except as regards the profitable disposal of the residuals of their undertakings. Mr. F. Livesey, of the South Metropolitan Gas Company, long ago reported most favorably on the utility of the appliance, and I am in