May 12, 1892] 



NATURE 



43 



to I '73 lb. of oil. At the next year's show the consumption 

 fell to I -42 lb. ; in 1890, to 1-243 'b., and Prof. Unwin this 

 year reports that a brake horse-power has been obtained by the 

 combustion of 0946 lb. Much yet remains to be done. The 

 useful work on the brake is under 14 per cent, of the energy 

 latent in the fuel, while the heat carried off by the water jacket 

 round the cylinder, and by the exhaust is equivalent to 75 per 

 cent, of the total thermal capacity of the fuel. Dr. Anderson 

 was of opinion that a combination of the direct combustion 

 engine with the spirit-engine of the Yarrow type would give the 

 best results, especially if a more advantageous cycle than that 

 of the Otto gas-engine can be adopted. 



The address next proceeded to deal with the question of the 

 capacity of the earth to supply the ever- increasing demand for 

 petroleum, and to enquire whether it would be possible to 

 substitute it largely for coal as a source of heat, owing to the 

 fact that we should have to go deeper and deeper in the future 

 to reach workable coal measures. In connection with this 

 problem the address gave particulars of the researches of 

 Mendeleeff, and described his theory of the continuous forma- 

 tion of petroleum by the action of water on the molten rocks in 

 the interior of the earth. The speculation is one of great 

 interest, but has already been dealt with in these pages. 



The vote of thanks to the President for his address was 

 moved by Sir Frederick Bramwell and carried with accla- 

 mation. 



After the reading of the address the Report of the Marine 

 Engine Trials Research Committee, which had been prepared 

 by the Chairman of the Committee, Prof. Alexander B. W. 

 Kennedy, was read. This report dealt with the trials of the 

 Belgian channel steamer Villi tie Dottvres, which had been 

 generously placed at the disposal of the Committee by the Bel- 

 gian Government. This vessel is one of the line which carries 

 the mails between Ostend and Dover, and was built andengined 

 by the Societe Cockerill, of Seraing, Belgium, and is a com- 

 paratively new vessel, having been delivered in the year 1890. 

 The propelling machinery consists of a pair of compound sur- 

 face-condensing piddle engines. Vessels of this class are mainly 

 designed with a view to speed, as the chief object desired is to 

 carry passengers and mails quickly from port to port. As the 

 run is only of three hours' duration, it would obviously not pay 

 to enter into any refinements with a view of economizing fuel. 

 The time under way is comparatively small when considered 

 in relation to the time spent in raising steam and cooling down 

 again. This is a point which should be borne in mind, but 

 which some critics appeared to forget during the discussion. 

 Perhaps engineers are apt to base their estimates of efficiency, 

 especially in marine practice, too much on an economy 

 basis. It is a good thing to save fuel if it can be 

 done without too much sacrifice. An examination of 

 the details of the various trials of steamships made by the 

 Research Committee illustrates this important point. We hardly 

 know how to deal with this paper. It is full of information of 

 the most valuable description, but its very fullness renders it 

 extremely difficult to make an abstract, and we have not space 

 to give all the details in full. Perhaps the best plan will be to 

 give some of the leading facts ; and, although these may appear 

 somewhat bald standing alone, they will enable our readers to 

 form an estimate of the scope of the trials, and those who are 

 especially interested will go to the original, in the Transactions of 

 the Institution, for fuller details. The Ville de Doiivres is 271 

 feet long, 29 feet broad, and 15 '5 feet deep, moulded. Her 

 registered tonnage is 855 gross ; and her displacement 1090 

 tons. She was run for nine hours especially for the trial in the 

 North Sea. The engines are of the compound inclined, surface- 

 condensing type, with cylinders 50"I2 inches and 97'I2 inches 

 in diameter, with 72 inches stroke. Neither cylinder is steam- 

 jacketed, but there is an intermediate receiver encircling the 

 high-pressure cylinder ; an arrangement which certainly does 

 not tend towards efficiency. The air, feed, and bilge pumps are 

 driven from the main engines. The circulating pump is 

 separate, and is estimated to develop 47 indicated horse-power. 

 The surface condenser contains 6540 square feet of tube 

 surface, and it is so arranged that the circulating water passes 

 three times through the condenser. The course of the water 

 is such that the coldest water meets the hottest steam. This is 

 naturally not the best arrangement, for the circulating water 

 would still be efficient for taking heat from the hottest steam, 

 even after it had been somewhat raised in temperature by the 

 coldest steam. On the other hand, circulating water having 



been heated by the steam at highest temperature, will be com- 

 paratively inefficient to further cool down steam already cooled 

 to a great extent. In any case, if a good vacuum be ulti- 

 mately obtained, the refrigerating surface will be far less 

 effective. The paddle-wheels are 22 feet 10 inches over the 

 floats, the latter being 10 feet broad and 4 feet 4 inches deep. The 

 immersion on trial was 1 7 inches. There are four single-ended 

 return-tube boilers, 13 feet by 10 feet. The grate area is 236 

 square feet, and the total heating surface 7340 square feet. 

 There is forced draught on the closed stokehold system. The 

 total weight of all machinery, exclusive of paddle-wheels, and all 

 water is 361 tons. Block fuel was used thro-aghout the trial. The 

 calorific value, calculated from analyses made, was 14,390 

 thermal units per pound. This corresponds to an evaporation 

 of 14-90 pounds of water from and at 212° F. A number of 

 samples of furnace gases were collected and analyzed, with the 

 following mean results : — 



Oxygen. Nitrogen. 

 . 7-95 -. 8050 



. 8-44 ... 74-72 



There was a little uncertainty about the temperature of the 

 chimney gases, but the mean temperature wai assu ned to be 

 910° F. The mean draught was equal to a pressure of from 

 0-92 to I "22 inches on the water-gauge. .\ notable feature about 

 these trials was that the feed measurement was made by meters. 

 This is a vast improvement, in one respect at least, and that of 

 great importance, on the measuring tank system. Measuring tanks 

 are always cumbersome and difficult to fit ; so much so that they 

 generally prove the greatest bar to proper trials being made of the 

 efficiency of marine machinery. The meters used were of the 

 Kennedy type, and appear to have answered the purpose admir- 

 ably. There is no trouble in taking a meter reading, whilst the 

 measuring tanks require constant attention. We look on the 

 introduction of the water meter for this purpose as a most im- 

 portant step in advance, and one which will lead to engineers 

 obtaining more frequent mformation on the efficiency of marine 

 engines. It is most desirable that the performance of the boiler 

 should be separated from that of the engine. The indicated 

 horse-power and coal consumption give the economy of the 

 whole machine ; but when results are not satisfactory it is often 

 difficult to say whether the fault rests in the boiler compartment 

 or the engine-room. Another step in advance is the effort made 

 to measure the amount of priming water. In the present day 

 we do not have so much trouble from priming as in past times, 

 when lower pressures were in use and the steam space was 

 practically what it is now. Still, there are yet large quantities 

 of unevaporated water often carried over to the engines by the 

 rush of steam. It is obviously useless to exercise great care in 

 measuring the feed if a considerable part of it is carried 

 from the boiler to the condenser simply as water. In 

 such a case the boiler is credited with a high evaporative 

 efficiency by reason of its very fault ; and the engine is debited 

 with steam which it never receives, but on the contrary is having 

 its action impaired by the presence of water in the cylinders. 

 The method of testing for priming is as follows : — A quantity 

 of steam from the main steam-pipe is condensed in a special 

 surface-condensing apparatus, and collected, and at the same 

 time a sample of water is taken separately from the boilers. 

 Both of these samples are carefully analyzed to determine the 

 quantity of salt present in each. As the whole of the salt found 

 in the sample from the steam-pipe must have come over from the 

 boiler in conjunction with priming water, and not with steam, 

 a simple calculation will show how much boiler water corre- 

 sponds with the quantity of salt, if any, found in the steam-pipe 

 sample. From this it is easy to determine what percentage of 

 the whole feed-water has passed from the boilers in the form of 

 water, or, in other words, what percentage there is of priming. 

 The chemical determination for salt is a very simple one, and is 

 capable of being carried with ease to an exceptional degree of 

 certainty. The observed and calculated data of the trial are given 

 in a full table appended to the report. The mean boiler pressure 

 was 105-8 lbs. above atmosphere, the vacuum 10-12 lbs. below 

 atmosphere, the revolutions 36-82 per minute, the mean indicated 

 horse-power 2977, the fuel per square foot of grate per hour 

 31-3 lbs., and the feed-water per indicated horse- power per hour 

 20-77 lbs., allowing for auxiliary engines. The efficiency of the 

 boiler was 66-1 per cent., and of the engines 1 1 ^^ per cent. The 

 combined efficiency of engine and boilers was 7*7 per cent. 



NO. II 76, VOL. 46] 



