Mr. Thornycroft has for some years used the locomotive form of boiler for his steam launches, working them under an air pressure--produced by a fan discharging into a close stokehold--of from 1 in. to 6 in. of water, as may be required. The experiments made gave an evaporation of 7.61 lb. of water from 1 lb. of coal at 212° Fahr., with 2 in. of water pressure, and 6.41 lb. with 6 in. of pressure. These results are low, but it is to be remembered that the heating surface is necessarily small, in order to save weight, and the temperature of the funnel consequently high, ranging from 1,073° at the first pressure, and 1,444° at the 6 in. With the ordinary proportions of locomotive practice the efficiency can be made equal to the best marine boiler when working under the water pressure usual in locomotives, say from 3 in. to 4 in., including funnel draught.

It has fallen to the lot of the writer to fit three vessels recently with boilers worked under pressure in closed stokeholds. The results, even under unfavorable conditions, were very satisfactory. The pressure of air would be represented by 2 in. of water, and the indicated horse power given out by the engines was 2,800, as against 1,875 when working by natural draught, or exactly 50 per cent. gain in power developed.

Mr. Marshall then proceeded to refute the arguments which may be urged against the use of the locomotive boiler at sea, and which we need not reproduce. Coming to the engines, Mr. Marshall said that the total working pressure of to-day may be accepted as 105 lb., or equal to seven atmospheres. If it were boldly accepted that eleven atmospheres, or 165 lb., were to be the standard working pressure, the result would be a gain of 14.55 per cent., provided no counteracting influence came into play. Of course, there are forces which war against the attainment of the full extent of this advantage, viz., the greater condensation in the cylinders and loss in the receiver or passages.

In regard to the former, it may be questioned whether by steamjacketing the high pressure cylinder, correctly proportioning the steam passages, and giving a due amount of compression in both cylinders, this may not be reduced far below the generally received notion; and the latter cause of loss may be considerably reduced in its effect by a more carefully chosen cylinder ratio. The ratio usually adopted, between 3.5 and 4 to 1, whether the pressure be 70 lb. or 90 lb., may well be questioned. With a cylinder ratio of 2.95 to 1, the economic performance is very good, and equal to any with the higher ratio. A lower cylinder ratio has another advantage of considerable value, viz., that the working pressure can be much reduced as the boilers get older, while by giving a greater amount of steam the power may be maintained--at an extra cost of steam, of course, but not so great a cost as with higher ratios. The cut-off in the high-pressure cylinder usually takes place at about 0.6, and the ratio of expansion has decided the ratio of cylinders. The use of separate starting valves in both cylinders obviates that necessity.

The difficulties in the way of taking advantage of the higher economic properties of greater pressures than hitherto used on board ship, are, it is submitted, not insuperable, and it would be to the interest of all that they should be firmly and determinedly met. It may be accepted as an average result that the Woolf engine, as usually arranged, will use 10 per cent. more steam than the receiver engine for the same power.

Of the three-cylinder receiver type the data are insufficient to form a definite opinion upon; but so far the general working of the Arizona is stated to be as good, economically, as any of the two-cylinder receiver class. The surface condenser remains as it was ten years ago, with scarcely a detail altered. In most engines it remains a portion of the framing, and as such adds greatly to the weight of the engine.

It is a question seriously worth consideration whether or no the surface of tubes can be reduced. The practice at present is to make the surface one-half the boiler surface as a minimum, that is, equal to about 2 square feet per indicated horse power. In practice, the writer has found 1.4 square feet per indicated horse power to maintain a steady vacuum of 27½ inches.

Mr. Marshall has just completed six pairs of engines for three twin screw ships, having steel shafts of 10 inches diameter, and has in each case run the engines at 120 revolutions per minute, while indicating 1,380 horse power from each pair for ten to fifteen hours without stopping; and in no case has a single bearing or crank pin warmed or had water applied, the surfaces on examination being perfect. In these engines all working bolts, pins, and rods, except the piston and connecting rods, are of steel, all rods in tension being loaded to 8,000 lb. per square inch. The boilers are of the Navy type, made throughout of Siemens-Martin steel plates, riveted with steel rivets, all holes drilled. Furnaces are welded and flanged; the tubes are of brass. In comparison with an ordinary merchant steamer's iron boilers of the double ended type, they weigh, including water and all appurtenances, as follows:

 Double ended Type. Navy Type. 
Weight, tons............ 135 ........... 146 I. H. P................. 1,400 .......... 2,760 Draught................ Natural ......... Forced.