in such a way as to form "box-girders; " while the outer sides and ends are ordinary web-plate girders. The bottoms of the purifiers are composed of wrought-iron or steel plates, slightly dished for the purpose of drainage, and for reducing strains due to gas pressure. It is also proposed to stiffen up the sieve-bearing bars in combination with the general structure of the purifier, so that they collectively act as girders and struts for carrying the weight of material, and to prevent distortion of the girder sides. The purifiers are shown to be supported by short columns placed directly under the boxes; but they could be equally well carried on the outer sides, by prolonging the girders on to the walls of a building the central columns only being retained.

DIAGRAM 3.

Each of these

The spent purifying material can be readily discharged through the sectional discharge-outlet shown in No. 1. sections should be made light, so that it can be easily handled by two men, and provided with a padlocked discharging door. It is also possible, with the system shown, to materially reduce the weight of pipe connections; and as these represent in ordinary practice from 9 to 30 per cent. of the weight of the purifiers and covers, it is evident that there is room for economy. By the aid of hydraulic valves formed in the body of the box-girders, as shown in No. 4, and also by utilizing the space in the interior of the box-girders as gas-passages, it is quite practicable-assuming proper workmanship in erection-to work a set of four purifiers

on a rotation system, with a less extent of pipe connections than is necessary with an eight-way centre-valve.

The use of some system of reinforced concrete flooring, such as the Monier system, or the "expanded" metal form of construction for the revivifying floor, is recommended. Full particulars of the latter system are given in a paper by Mr. A. Baker, read before the Institution of Gas Engineers in May, 1900; and although it has only recently been introduced, it has already been employed at Beckton and other gas-works for a similar purpose. The revivifying floor is shown to be enclosed by cast-iron parapet plates,

DIAGRAM 4.

DIAGRAM 5.

with a light single lattice girder secured to the upper flanges, for the purpose of carrying the roof principals. It is self-evident that the weight of metal in overhead purifiers constructed as described must be very much less than in any water-lute system. It must be borne in mind, however, that cast-iron plates cost only about one-half that of wrought-iron or steel work, such as is shown. The degree of ultimate economy likely to be attained depends entirely on the skill of the individual designer; but the author is of opinion that the purifier-boxes and covers alone can be

erected for about the same cost as those of equal capacity on the ordinary water-lute system, and that the cost of the supporting structure and connections can be reduced by an average of 50 per cent.

The author regrets his inability to give a greater amount of detail in the diagrams owing to the time for the preparation of his communication being somewhat limited.

The most complete and exhaustive system of sulphur purification practised almost exclusively to meet the regulations prevailing in London, remains in very much the same condition as when it was evolved many years ago. The author ventures to assert that very few of those who are, or have been, engaged in the practical working of the system will claim that it is an entire and economical success, and more especially under present conditions, when the gas manufacturer has perforce to be content with the supply of "all sorts and conditions" of a mixture called coal, giving a low yield of gas and residuals, but a very high one of other things, not the least of which are sulphur compounds and sulphuretted hydrogen. The introduction of air in purification has been of undoubted value in those happy places where sulphur clauses are unknown; but it is an unfortunate fact that it cannot be employed to the same extent under more rigorous conditions.

The hopes of the possibility of the purification of gas in closed vessels, raised by the advent of the Claus process, still remain unfulfilled; and the "process" itself has apparently now been relegated to the position of a thing of the past. The principle underlying the process, however—that is, "self-purification"— still remains, and is worthy of attention. A study of the very able and candid account of the failure of the experiments on the Claus process at Belfast, given in a paper read by Mr. Stelfox before the Incorporated Gas Institute in 1897, affords a good clue to the defects in the mode of application of the underlying principle. The non-success of the experiments appears to have been quite as much due to mechanical and constructional defects as to chemical ones. There is no reason why the former could not be overcome at the present day; thus restricting the problem to one requiring chemical solution only. The latter is again reduced by the proved fact that the process was found to be quite capable of effecting complete purification for consecutive periods of two and three weeks.

The reason why the system was subject to fits of bad behaviour has still to be solved, and deserves our attention. A perusal of the papers dealing with certain phases of this subject by

Mr. George Livesey and Mr. F. D. Marshall, read before the Gas Institute in 1875 and 1878 respectively, sheds some light upon the phenomena. The author is of opinion that if some scheme of experiments could be initiated under the direction of a Committee of practical men, a solution of the question would very soon be discovered.

Hills's liquor purification process in use at the Old Kent Road Gas-Works for many years, and the more recently adopted Holgate's liquor purification system at Halifax, have proved successful within the limits of their application. Need we, therefore, despair of the devising of some intermediate scheme between the extreme pretensions of the Claus process, and the more modest aspirations of the two latter?

It may not be out of place, in conclusion, to express regret that very little progress can be claimed for improvements in the general process of purification.

Discussion.

Mr. S. Y. SHOUBRIDGE (Lower Sydenham) said he was sure the Association were much indebted to Mr. Lewis for his interesting and suggestive paper, and all the more so because it was prepared at the eleventh hour, in response to an appeal from the Hon. Secretary. The chief feature of Mr. Lewis's design appeared to him (Mr. Shoubridge) to be the combination of the purifiers with the purifier-house; and though there seemed to him to be some doubt as to the wisdom of making purifier-boxes of steel or wrought iron, owing to the greater liability of corrosion, there was certainly very great advantage in being able to save such a large amount in the cost of construction as Mr. Lewis found that he could do by this method. Fifty per cent. in the cost of construction was a considerable saving. But he (the speaker) thought it was very important that the purifier itself should be kept quite distinct from the superstructure, because there was no doubt the purifier itself would wear out considerably before the other part of the structure; and in that case, if it had to be renewed, it might interfere with the stability of the building. There was, however, also a considerable saving in space; and that was a matter of importance in very many works-especially in those where they were at all cramped. Beyond this, the system was similar to that which had been in use in a great many gas-works. At Lower Sydenham, they had a purifier-house very much like the one shown in the diagrams, in which the indiarubber lutes were employed; and the purifying material was shot down in shoots like those described in the paper. They