croft. Mr. Thorny- of fire due to the quantity of incombustible matter in it at the end of the trial was a matter of no consequence, and did not affect the accuracy of the result obtained. There was, perhaps, some force in the objection that the 1.9 per cent. loss by radiation in trial D was too small; but it was also true that the loss due to furnace gases was more difficult to measure than the heat given to the steam. He would therefore claim that it was at least equally fair to take it from the one as from the other. It was observed that there was no loss due to smoke in trial D, but there was some loss from this cause in the trials at higher speed of working. He thought Professor Unwin had put too liberal a value on the possibility of error in the experiment considered. Mr. Spencer saw difficulty in firing so large a grate as was used in the experiments. This difficulty was imagined by considering a fire in an ordinary flue, where there was not room to throw the coal without striking the top of the furnace; a longer fire could be worked with advantage in a high furnace than could be managed in a low one, and extra width did not seem to affect the matter, except to give better or more complete combustion. He was glad to have Mr. Spencer's support on this point. In answer to Mr. Cowper, he thought large size of tube was not favourable to success, except when discharging below water-mark. The cause of failure in large tubes had been given in the Paper. To get high duty from the fuel required attention to the fire at frequent intervals, in any kind of boiler. Admiral Selwyn had made some remarks, to the effect that the tubes were only retained in the tube-plates by ferrules and ordinary expansion. As a matter of fact, no ferrules were used, expansion in the tube-plate affording ample security for a good joint and freedom from danger of the tubes being blown out. He considered that in using thin tubes, if the ends could be threaded and screwed in, the reduction of strength due to the thread would more than compensate for any advantage that could be gained. Mr. Stroudley had directed attention to the fact that in the blacksmith's tuyere a similar method of circulation as that adopted by the Author had long been employed, and the tube used was not blocked by incrustation, although exposed to a very intense and local heat. The only difficulty in this was that of making repairs in the present form of boiler. This difficulty, he thought, was not so great as had been suggested by several speakers. If a tube failed at the inner part of the system, several other tubes would have to be removed before it could be got at; but these tubes could be removed and replaced again with little difficulty; being long in proportion to their diameter their curved form gave them great elasticity, and the ends of a tube could be moved Mr. Thornyrelatively to each other through a considerable distance without croft. causing any permanent bend. This enabled them to be adjusted in their place although not quite accurately bent. He did not consider Mr. Hodge's description of the boiler gave a correct idea. In what were known as vertical boilers, the gases usually escaped from the upper part of the firebox, while in the Thornycroft boiler there was a large volume in the firebox above any point where the gases could escape, and the heating-surface was not vertical, as described by Mr. Hodge, but inclined. Mr. Duckham was anxious to accept the high evaporation obtained by the boiler, while he at the same time took away the means by which this was obtained. The tubes were not curved in the particular way in which he found them, to allow for the difference of expansion between those which were more particularly under the influence of heat, and those which were less heated, but were curved so as most conveniently to form a firebox and flues, while, at the same time, the separator was protected from heat. This, he believed, could not be done with straight tubes, even if otherwise unobjectionable. Mr. Willans wished to know if any statistics could be produced as to the efficiency of the separator. He had not tried any experiments to determine this point; but the working of the engines tended to show that less water came over than with the locomotive boilers previously used, and he thought what Professor Kennedy had said was conclusive as to the comparative dryness of the steam in trial D, where the highest evaporation was obtained. When the boiler was worked very hard, sometimes a little priming was evident; but in the experiments water had been kept especially low in the separator, with a view to prevent this; and the fact that all the heat was accounted for seemed to indicate there could not be much water passing over with the steam. Mr. Willans had taken the consumption of steam by the engines as an indication that the steam was wet; but it had already been explained that the proper feed-pumps delivering into a tank in the stoke-hold, all the feed-water had to be pumped a second time, and against the full boiler-pressure, by a donkeypump, in such bad order, that when used intermittently on the first trial, the amount of steam taken was so great that a fall of pressure could be at once detected when it was set in motion. But to settle this question of dryness of steam he hoped to make some experiments, the result of which he should have pleasure in making known at some future time. In examining the calculations made by Mr. Halpin, he found that a trial of the water-tube boiler croft. Mr. Thorny- had been taken when the rate of working was not so rapid as in the trial of the locomotive with which it was compared. If he had taken trial B, where the economy of evaporation was more nearly the same, he would have obtained 66 as the value of the factor. Instead of being 9 per cent. less than that obtained from the locomotive, it was 32 per cent. greater. For trial E, that at the highest rate of working, it would be above 80. Mr. List had discussed the suitability of the boiler for ocean steaming, and it appeared probable that a field was open here where high speed or very long voyages rendered saving in coal important. Unfortunately, information about the performance of marine boilers was at present not sufficiently complete to enable him to give an exact comparison. In the figures given by Mr. List, the boiler efficiency was obscured by being taken together with that of the engines. The great weight, however, of the boilers was evident; and the fact that a much larger heating-surface could be given in the same space showed that the ordinary boilers brought forward by Mr. List might be displaced with considerable advantage, both in the decreased weight of boilers and weight of coal to be carried. Mr. List believed that the only danger to be feared with the Thornycroft boiler at sea would be from burning the tubes if the condenser leaked considerably, or in the event of greater leakage than could be made up by the evaporators. In order to avoid leakage of condenser tubes, Mr. Normand of Havre had modified the form of his condenser, making all the tubes curved, the segment of a large circle. This construction did not appear to involve any serious difficulties, and Mr. Normand had informed Mr. Thornycroft that these curved tubes, simply expanded into the plate, kept perfectly tight, and were not liable to leak after being out of use a short time, as was the case with the ordinary condenser. This construction evidently formed a solution for one of Mr. List's difficulties. In his opinion Mr. John Head was not justified in comparing the combustion of mixed gases with the combustion of solid fuel on the same terms. Although in the experiments a certain amount of free oxygen passed up the funnel, in most of them the result would have been better had this quantity been greater. The results obtained by Mr. Head in firing boilers by producer gas were, as he suggested, not very favourable, and this was probably due to the causes named. He was of the opinion, however, that if the fuel in a ship was to be turned into gas, it might perhaps be burnt in the cylinder of the engine with advantage. This view was held by the late Sir William Siemens; but if a steam-boiler must be used, it would no doubt be most convenient croft. to regulate all the fires by simply opening or closing a valve. It Mr. Thornywould have been interesting if Mr. Head would have mentioned in how small a space, and of what weight, the gas producer could be made for some particular amount of evaporation. Correspondence. Mr. W. INGLIS said he had taken a good deal of interest at one Mr. Inglis. time in this kind of boiler, especially as regarded the circulation through the tubes, having designed two types with this feature particularly in view, in 1863, and several had been made about that time by Messrs. Sir W. G. Armstrong and Co. and other firms. He had also read a short Paper on water-tube boilers at the meeting of the British Association for the Advancement of Science held at Newcastle-on-Tyne in that year, when he exhibited a working model of his water-tube boiler with the tubes made in glass, to show the perfect circulation of the water through them. The advantage of this circulation of the water had been, he thought, satisfactorily proved from some of the boilers he 'made about 1864 having been in constant use night and day at papermills near Edinburgh, where, two years ago, he saw them still at work, and he was informed that two of the boilers had been in use till April last, when they were taken out after working for twenty-five years. Each of them was equal in capacity or steaming-power to an ordinary Lancashire boiler 7 feet in diameter and 30 feet long. He thought the Author was right in aiming at having a very free circulation of the water. In nearly all the water-tube boilers brought out in recent years, arrangements for securing or promoting circulation had been more or less imperfect or neglected altogether. Mr. W. KILVINGTON believed there was an impression that water- Mr.Kilvington. tube boilers were more economical of fuel than smoke-tube boilers. But there was no reason so far as he could see why this should be · so; and he did not know of any experiments which proved it. He could not think why a smoke-tube boiler should not be as economical as any other, because good combustion could be got in properly-proportioned furnaces, the gases in the chimney could be kept at a moderate temperature, and, as the heat could not from the construction of the boiler be lost in any considerable quantity by radiation, it must have gone into the water and so raised steam. The circulation might not be so good and Mr.Kilvington. the steam-space might have to be larger. But in viewing boilers generally, he thought it would be admitted that the passage of the products of combustion from the fires to the uptake, in such a way that they were brought in contact with all the heating-surface, was quite as important as the circulation of the water inside, and this was provided for in the ordinary marine type of boiler more surely than was the case with most water-tube boilers. It did not appear that the experiments of Professor Kennedy exhibited any economy in the Thornycroft boilers over good examples of the usual marine type; for, although series D, Table I, showed a very large efficiency, namely, 86.8 per cent., yet, when the easy conditions of the trial were taken into account, it was not so extraordinary, for the boiler had more than 9 square feet of heating-surface to absorb the heat from each lb. of coal burnt per hour. He thought an equal efficiency might be predicted for an ordinary smoke-tube boiler of the common marine type worked under the same easy conditions, and this assumption appeared the more probable on examining series B, Table I. It would be seen that here the quantity of water evaporated from the temperature of the feed was 9 6 lbs., and the heating-surface was a little more than 2 square feet for each lb. of coal burnt per hour. This was about the ratio of heating-surface with which hundreds of marine boilers worked at sea, which certainly evaporated 10 lbs. and upwards of water per lb. of fuel, so that their efficiency would exceed that of the boiler under consideration. Boilers of the water-tube type were thought to be a necessity when high pressures were first mooted for use at sea; but now boilers of exactly the same type as were used for pressures of 60 to 80 lbs. were daily made to work at 150 to 170 lbs. per square inch, and were in every way successful. Undoubtedly the boiler under discussion was an excellent one, but still he was firmly persuaded that the time was far distant when water-tube boilers would be employed at sea, except in cases where extreme lightness was imperative. Mr. Kirk. Mr. A. C. KIRK remarked that the boilers of the "Propontis" had been designed by Mr. Rowan, and were, he understood, practically duplicates of boilers previously worked with success on the "Ganges." It was to utilize the exceptionally high pressure of steam afforded by them that he designed the three-crank tripleexpansion engines; and it was the failure of these boilers that prevented him from repeating them till about eleven years later, when they were adopted in the s.s. "Aberdeen." Soon after, this type of engine became almost universal. The failure of Mr. Rowan's |