Comparing naphtha refuse and anthracite, the former has a theoretical evaporative power of 16.2 lb. of water per lb. of fuel, and the latter of 12.2 lb., at a pressure of 8 atm. or 120 lb. per square inch; hence petroleum has, weight for weight, 33 per cent. higher evaporative value than anthracite. Now in locomotive practice a mean evaporation of from 7 lb. to 7½ lb. of water per lb. of anthracite is about what is generally obtained, thus giving about 60 per cent. efficiency, while 40 per cent. of the heating power is unavoidably lost. But with petroleum an evaporation of 12.25 lb. is practically obtained, giving 12.25/16.2 = 75 per cent. efficiency. Thus in the first place petroleum is theoretically 33 per cent. superior to anthracite in evaporative power; and secondly, its useful effect is 25 per cent. greater, being 75 percent. instead of 60 percent.; while, thirdly, weight for weight, the practical evaporative value of petroleum must be reckoned as at least from (12.25 - 7.50)/7.50 = 63 per cent. to (12.25 - 7.00)/7.00 = 75 per cent. higher than that of anthracite. _Spray injector._--Steam not superheated, being the most convenient for injecting the spray of liquid fuel into the furnace, it remains to be proved how far superheated steam or compressed air is really superior to ordinary saturated steam, taken from the highest point inside the boiler by a special internal pipe. In using several systems of spray injectors for locomotives, the author invariably noticed the impossibility of preventing leakage of tubes, accumulation of soot, and inequality of heating of the fire box. The work of a locomotive boiler is very different from that of a marine or stationary boiler, owing to the frequent changes of gradient on the line, and the frequent stoppages at stations. These conditions render firing with petroleum very difficult; and were it not for the part played by properly arranged brickwork inside the fire box, the spray jet alone would be quite inadequate. Hitherto the efforts of engineers have been mainly directed toward arriving at the best kind of "spray injector," for so minutely subdividing a jet of petroleum into a fine spray, by the aid of steam or compressed air, as to render it inflammable and of easy ignition. For this object nearly all the known spray injectors have very long and narrow orifices for petroleum as well as for steam; the width of the orifices does not exceed from ½ mm. to 2 mm. or 0.02 in. to 0.08 in., and in many instances is capable of adjustment. With such narrow orifices it is clear that any small solid particles which may find their way into the spray injector along with the petroleum will foul the nozzle and check the fire. Hence in many of the steamboats on the Caspian Sea, although a single spray injector suffices for one furnace, two are used, in order that when one gets fouled the other may still work; but, of course, the fouled orifices require incessant cleaning out.