The Chairman said that the Author had given them a very interesting paper, the more interesting as it was the result of his own experiences, and especially in connection with the articulated locomotive, a design novel to us on this side of the water. He did not think it at all likely that we should see Mallet locomotives on this side for several reasons. In the first place the turntables would put them quite out of the question, and in the second place unless we could make use of the great adhesive weight given by the Mallet locomotive it would not be of much use to us. If we use a Mallet it means we get at least six working axles, which adds to the adhesive weight, and to make use of it we must have larger cylinders which mean larger boilers, and in that direction we are limited by our load gauge. The Canadian and American load gauges are much more generous than ours which is now practically up to the limit, so unless we can find souse way of increasing the load gauge we cannot make use of the greater adhesive weight of the Mallet locomotive.

He said there was a point mentioned about the control of tho side movement of the bogie. The inclined plane arrangement which is similar to that used on the G.W.R. pony trucks was supplemented by the use of two springs which acted partly as control and partly as a centering device. He would like to ask if there was any initial compression put in these springs when the pony is in its central position. The Author replied that this was so, there was sufficient initial pressure to bring the initial rolling control to the figure he had mentioned. The roller travelled flat for two inches off the central line of the boiler, and the screws were compressed so that the resulting pressure was about 1,650 pounds. Another point in comparing the Canadian with ours was the allowable weight per axle. For the Pacific type of engine, the load per axle worked out at about 25 tons, which figure was quite out of the question with us on account of the permanent way, but for the other types of engines the load per axle was in the region of 20 tons.

The author said he did not suggest that Mallet locomotives should be used in Great Britain. They could not, economically, because the only type we could use here would be a small one which could be turned on a 70 foot turntable, and small Mallets produced usually excessive maintenance costs, which really went to show that for a Mallet engine to pay the weights which it has to haul must be far and above that which any ordinary type can handle. Such big engines as this would be impracticable in this country. Also a 2-10-0 engine can pull practically an equal weight to a Mallet engine of the same weight, and with less the maintenance cost.

Mr. Stanier said that he remembered early in 1901, when the new shop was being built, (Swindon Works) that it was reckoned to be the last word in erecting shops. The late Mr. Dean designed it to deal with the largest engines he could conceive would run on English railways, i.e., the 2600 class and the "Cities". Within six years the "Great Bear" was built, and that is still ahead of our own Engineering Department. Twenty tons per axle is the limit for modern bridges on the G.W.R. (Great Western Railway), and the "Great Bear" has 20 tons on an axle. That was 15 years ago, so that in this country, although we are confined to such small engines, the Locomotive Department is generations ahead of the Engineering Department. Engine No. 4,700, one of the latest developments on the G.W.R., has 19 x 30 cylinders, and a boiler that will put say 18 or 18 1/2 tons on each axle. That engine will be able to run on any main line. He said one is interested to see that in Australia they can develop a 2-10-2 with their advantages of load gauge, and the risks their Engineering Department is prepared to take. One could only take it to be risks when one remembers the trestle bridges they ran over, and a 2-10-2 with enormous boiler and cylinders cannot at any rate be a cheaper engine to run. He said the Author had not said anything of the efficiency of the engines. He had compared the efficiency ratios on some of the American and G.W.R. engines. Some years ago an American Company built an engine that was to be the last word in locomotive construction. On comparing this with the "Great Bear" ratios, we were very much ahead of the last word in American locomotive design. Could the Author tell us how the efficiency ratios of these Canadian Pacific engines compare with modern English Locomotive practice? Modern English design is as efficient as experience can make it. The American special engines are built of alloy steel parts, and the piston rods are hollow and heat treated. The connecting rods are nickel chrome and heat treated, yet the efficiencies are lower comparing the American engine with the "Great Bear". He said the G.W.R. had used a great deal of alloy steel experimentally, but the general practice was to use straight carbon steel. He would be interested to hear if the author had any information to give as to whether alloy steel was found advantageous in these big engines.

Replying to Mr. Stanier, the author said there were now no wooden bridges on the Canadian Pacific Railway. When the first urgent demands came for increased power on the Rocky Mountain section there were, and that was one reason why the weights had to be kept down. Then the engines had to keep to 19 to 20 tons per axle (similar to over here), including the 0-6-6-0 Mallet described. The trestles have all since been replaced by steel structures, and the Engineering Department now allows large engines to run carrying about 25 tons per axle. With regard to the latter part of Mr. Stanier's inquiry, he could not say very much as it was six years since he was in Canada. Many of the figures he had given in the paper were sent to him by the Chief Mechanical Engineer of the C.P.R., and they were not sufficient to work out all the efficiencies. He did not think the vast distances of the North American Continent were appreciated when discussing English and American railway practice. The main thing over there is to get the trains along over immensely long sections. Railways were laid very cheaply in the first instance because they lad to get an enormous country developed, and frequently the townships would spring up almost alongside the laying of the track. The speeds run over there are not generally so high as on this side, such high speeds are not needed, because people do not expect to do business in two or more large cities in the same day as in England. He said the Chief Mechanical Engineer of the C.P.R. had kindly said that if the data sent was not sufficient he would be very pleased to supply additional data. The author would write and ask for the complete efficiency ratios. He said that when they built the Mallet the C.P.R. certainly did use alloy steel, but could not say what their 1921 practice was in this particular respect.

Mr. E.J.F. Plaister raised a question regarding flangeless regarding tyres and the arrangements for a long locomotive like the one described.

The author replied that he had recently obtained some information from Baldwin Locomotive Company about modern types of Mallet engines, and on these very long engines the designers usually leave off the flanges on a middle pair of wheels, but whether the Canadian Pacific do this on their 2-10-2 engines he could not say. He did not think there, was much difficulty in getting a 10 wheel-coupled engine to curve easily. They are low speed engines, and they have a traversing pony truck in front, and one at the rear. The speed they run is largely from 7 to 20 miles per hour, so there is nothing very dreadful, even in running round the sharp curves mentioned in the paper. One thing about Mallet engines is that there is obviously a point where a curve joins a straight, or runs from the straight, at which the articulated pin would be on the junction of the tangents of the curve, and this is the point of danger as the two trucks tend to separate, and the articulated pin has a snatch. The effective rigid wheel base, as far as the front truck of the 0-6-6-0 Mallet is concerned, is practically the distance between the rear flange of the front truck and the rear truck. The front truck guides from the articulated pin and not from the rear flange of the truck. In the centering arrangement on Mallet engines great care is taken to ensure that there may be a definite pull on the boiler through the medium of the front truck, the action of which on entering a curve was similar to that of an 0-6-0 tank engine. He could not say the thickness of the flanges through the 2-10-2 engines.

Mr. Hinton referred to the system of lubrication used on these locomotives, and said one was almost inclined to think there would be some trouble through the oil congealing owing to the low temperature. It was also mentioned in the paper that the pull-out regulator was standard practice. Exactly what advantage had that? He would also like to have some details of the equalizing pipes between the boiler proper and the feed water section. In the paper it was also mentioned that a very gentle blast was obtained. Did this do away with any need of a regulator on the blast pipe, such as a "jumper ring", which is standard G.W.R. practice?

Replying to Mr. Hinton, the author said that with regard to lubrication, so far as he knew, the C.P.R. used mechanical lubricators, generally speaking, and on the Rocky Mountain section he believed a lubricator somewhat similar to Mr. John Robinson's "Intensifore" was utilized, which kept the oil fluid in the cold atmosphere. He did not know whether there was any trouble through the congealing of the oil, but had no doubt but that the best system was the G.W.R. one, where the oil was fed into the steam pipe. With regard to the pull-out regulator, he said this was largely a matter of taste. A number of English locomotive engineers and an equal number of Americans would probably argue the whole evening on the subject and never get anywhere, but American and Canadian enginemen are accustomed to the use of a pull-out throttle. With regard to the equalizer pipes, he said these were 4 inch internal diameter. They were seated into saddles, riveted to the boiler shell, the seatings were conical, and the pipes held in place by bolts. He said these outside pipes might appear to be terrible things but no trouble was experienced with them. He said he remembered in 1906 Mr. Churchward reading a paper on large locomotive boilers before the Institute of Mechanical Engineers, in which he described some experiments made to ascertain the direction of water circulation, and which indicated that the water tended to travel from the front towards the fire-box. Mr. Churchward then suggested that it might probably be beneficial to the circulation if outside circulating pipes were fitted, and that such outside tubes had been used in France and America. With regard to the blast pipe, the author thought somebody would say that 4 3/4 inches seemed a small orifice for so large an engine. In comparing it with British and G.W.R. practice it does seem small, but it has to be realized that the 7 inch pipe leading from the L.P. cylinders to the blast pipe was a very long pipe and acted as a reservoir. The back pressure on the L.P. cylinders was small. There was no regulator on the blast pipe orifice, which was satisfactorily maintained at 4 3/4 inches right through the experiments.

Mr. Cook asked if the tests mentioned with the Mallet taking 700 tons were, as he presumed, single-handed or with a second engine in front, as it seemed an enormous strain on the couplings in view of the gradient. About what size was the coupling used? Also a question with regard to the fuel used. Was it the general practice to use coal? He saw a report some time ago that oil fuel had been used to a large extent.

With regard to the weight taken of 700 tons, the author said this was the maximum weight taken on that grade by the Mallet alone, but the normal weight was 640 tons which would allow the engine to work quite easily. So far as he knew, however, the Mallet was generally used as a pusher, in the rear of larger trains. The couplings used were the usual automatic couplings, standard for all rolling stock in America and Canada. They were very big things, massive castings which you could not break very easily.

With regard to fuel, he said that the latest types of locomotives built by the Canadian Pacific Railway were fitted to burn coal or oil. He did not know to what extent they are now using one or the other fuel.

Mr. W.A. Stanier said that the author had referred to coal or oil being used. He would like to know the author's opinion about the use of pulverized coal in the Rockies. He believed it was a very usual method of firing engines. We had also heard of some experiments with it in this country, but he understood that another method of using coal or oil is being experimented with. It is generally known as Colloidal Fuel, a mixture of oil and powdered coal.

The author replied that he could give no definite information with regard to powdered coal. All he knew of it was from a paper read by a gentlemen who advocated it, and that the G.C.R. found it a very dirty fuel to handle. As regards colloidal fuel, this was experimented with during the war, largely by the United States Committee of Submarine Defence. One of the American Torpedo Boats, the "Gem", used it with very satisfactory results, but the war came to an end before much had been done. Its claims are these. Varying percentages of liquid fuel, not necessary oil, residuals, even tar, can be mixed with varying quantities of ground coal (not coal dust, but ground in machines somewhat similar, he believed, to those paint is ground in) and these mixtures vary from fluidity through various viscosities to a thick paste. The mixture generally used is about 40 percent of ground coal and 60 percent of oil, and this makes a mixture which will flow. Its advocates, however, prefer to put a small rotary pump in the pipe line, which pump is really a displacer, and passes the fuel onto the fire-box. The pump is driven by a small steam turbine, and the driver can regulate the supply of fuel to a nicety. The fire-box nozzles used are exactly the same as for oil fuel. The advocates of colloidal fuel recommend that no difference be made in these. The principle claim for colloidal fuel is this, that if a low glade oil and a low grade coal are taken the resulting mixture is burnt with as complete combustion as oil fuel alone. However perfectly a fire-box may be designed a great number of potential B.T.U. from coal will go up the chimney. Combustion with straight oil fuel is far more complete, but oil is expensive. With colloidal fuel it is claimed that the combustion is equally good to that of straight oil, but that the cost per ton of the colloidal mixture is much less than that of oil alone, in other words, that colloidal fuel will give an equivalent or higher fire-box temperature per unit weight of fuel than with coal alone, at a cost, per heat unit absorbed, of less than that of either straight coal or straight oil.

Mr. Cuss asked how the colloidal fuel was conveyed to the fire-box, and Whether or not an ordinary coal-burning locomotive with an ordinary brick arch is suited for that purpose also for oil fuel, so that one fire-box constructed for one fuel would be suitable for liquid or solid fuel. Also when 2-10-2 locomotives are derailed, does it occupy the main lines for very long before they are on again, or have they special high-speed arrangements for replacing them at short notice? Another point, what arrangements are made for removing ashes and clinkers and cleaning the boiler tubes, and what arrangements have they whilst taking water to the boilers from the tenders to protect against the severe frosts.

Replying to Mr. Cuss, the author said that he thought it was impossible to burn three kinds of fuel efficiently in one box. The advocates of oil burning usually arrange that the fire-box shall be lined with fire-brick. The claim is that a locomotive fitted for oil can burn colloidal fuel without alteration. With regard to the displacement pump, the colloidal fuel comes down from the tender, passes under the fire-box and up about 10 inches under the boiler, feeding into the front of the fire-box. This flow is by gravity (assisted by the displacer pump), and the flame is towards the back of the fire-box. The jet advocated for colloidal fuel is not exactly a nozzle, but the fuel trickles out of a rectangular slot at the end, where it is met by the usual arrangement of steam jet and so carried into the fire-box. With regard to cleaning boiler tubes, the C.P.R. used to employ one of the common flexible arrangements. As regards clearing the ashes and cinders the arrangement was no different for the big engines than for the small ones.

Speaking of derailments, the author said he had never been in one on the C.P.R., but all engines carry re-railing ramps and can often pull themselves onto the rails again. There was no peculiar difficulty with big engines that he knew of.

Mr. W.A. Stanier said that a little time ago he was talking to two American engineers. They said with regard to the derailment of a large Mallet engine that it was quite simple. You put down the re-railing ramps, and the engines work themselves up them. The track out there is different to ours here. They can get a re-railing ramp that will work. It is difficult to get a re-railing ramp to fit a chaired road. He said they had experimented with many, but have never found one that is entirely satisfactory. It is difficult to get a ramp that will keep near the rail when the engine is going up. In the States, they use re-railing ramps, and sometimes large steam winches. It is a more simple matter to deal with derailments than in this country unless the engine overturns.

Mr. Manning asked if there was any reason for not attempting a higher boiler pressure than 200, also what type of injectors were used?

The author said that at the beginning of this century the C.P.R. started out with 200 pounds, and since then it has been standard. He said he believed that Mr. H.H. Vaughan, who was Superintendent of Motive Power till about 1915, saw a long way ahead of most other railways of the world with regard to superheating. Quoting a paper read before the Institute of Mechanical Engineers in the early years of this century, Mr. Vaughan had contributed some remarks to the discussion, saying that even then for some years every engine on the C.P.R. had been superheated, even the smaller ones. He started with a pressure of 200 pounds, and his successor has carried it on. There is no load gauge restriction for large cylinders, so the lower pressure can give all the power required. He did not know why they had not used a higher pressure. He said there had recently been articles in the "Railway Gazette" and elsewhere about a large 2-10-0 engine for the Great Indian Peninsula Railway (G.I.P.). That engine was given the low pressure of 160 pounds per square inch, and to develop the tractive effort necessary the designer had to resort to four cylinders, and correspondingly higher maintenance. It seemed to him that was the way not to do things. He did not know if any of them there know why they have such low boiler pressure in India. The G.I.P. engine referred to has four cylinders 20 inches x 26 inches stroke, and the tractive effort produced with 160 pounds boiler pressure is 53,000 pounds, the engine being built to full load gauge dimensions. If the minimum load gauge did not permit of two sufficiently large cylinders with this low pressure, to produce tractive effort aimed at, why not have raised the boiler pressure slightly? Had the boiler pressure been made the moderate one of even 200 pounds per square inch, two cylinders 23 inches x 32 inches stroke would have given an even higher tractive effort. After all 75 percent or more of the freight haulage of the world was being done by two cylindered locomotives.

The Chairman said that going in for four cylinders seemed to be a step in the right direction. The weight of the reciprocating parts for each cylinder is much smaller, and those parts balance in themselves, and so reduce the hammer blow on the rail, which in the case of a large two-cylinder engine is appreciably disturbing.

Mr. Stanier said it was largely a question of the speed the engine had to work at. If you have a big enough load gauge a 2-10-0 with 4 feet 8 inch wheels and two cylinders would never reach beyond 25 miles per hour, and the hammer blow is kept within limits. With regard to the low boiler pressure on the Indian Railways he said that in India they depended largely on native labour for maintaining the boilers, and they are afraid to risk putting native labour, which requires a large amount of supervision, on high pressure boilers above 160 pounds.

Mr. Kerry bore out Mr. Stanier's remarks regarding native labour in India. In Calcutta they had continual trouble with leaky stays due to the bad work put in by native workmen. You cannot depend on the work they do unless they are continuously supervised, and he said he thought the author would find that was the real reason of the low pressure boilers.

In closing the discussion the Chairman said the paper would be a useful addition to the Society's proceedings, and they would be glad to avail themselves of the author's offer to obtain the efficiency figures, to be bound up with the Transactions.