ARTICLE 2. Shorter cylinder toilers than most of those in common use will probably yet he employed. No man has devoted more attention to this question than C. Wye Williams, of Liverpool, England, the author of an excellent work on furnace and fuel, and the inventor of several improvements in boilers and furnaces. He has lately, through the columns of the London Engineer, published some very useful information regarding the evaporative effects of flue surfaces. A very general notion used to prevail in regard to the economy of long flues in tubular and cylinder boilers. It was held that the heat of combustion in passing through long flues or tubes gave out nearly all the caloric to the water, and thereby economized fuel. This notion is based upon the supposition that in short boilers the heat rapidly escapes, without being absorbed by the heating surfaces. Plausible as this idea is when first presented, it is easy, however, to perceive that a boiler may be made so long as to be incapable of generating steam for useful purposes. The concentrated heat generated at the furnace end of a long boiler may be carried off at a low temperature by a great mass of water, and thus becoming diffused, may be radiated from a great extent of surface, and fail to generate steam. Some have held that mere tube surface is equivalent to evaporative power, and nine square feet of such surface to one square foot of grate has been called " a horse power." Upon such ideas, many boilers, with very long tubes, and others with longer flues, have been made in the expectation that evaporative effects would be obtained in proportion to the extent of long flue and tube surface. Kecent experiments have demonstrated the inaccuracy of such notions. It has been found that heating surface rapidly diminishes in useful effect, as it recedes from the fire. The following figure is a longitudinal section of an experimental tubular boiler with which several experiments were made by John Dewrance, engineer of the Liverpool and Manchester Railway, England. This boiler was divided into six water compartments, in order to ascertain the weight of water evaporated in each. The first chamber was but six inches long, each of the others twelve inches. The tubes were five feet and a half long, and run from the fire to the smoke-box in the usaai way. In the first compartment, each square f bet of the heat absorbing surface was found to evaporate as much water as each square foot of the fire-box surface. In compartment 2, each square foot of heating surface was only about one-third as efficient as a like amount of surface in No. 1 in the remaining four compartments, the evaporation was so feeble, that Mr. Dewrance said, "I came to the conclusion that the first six inches of the tubular series had more evaporative effect tliiiri the remaining sixty inches." Similar experiments have also been made by C. Wye Wiliiams, the results of which have Ijeen published in the London Engineer. He employed a small cylindrical boiler divided first into five compartments, with a tubular three-inch flue running through the interior to the chimney. The first chamber near the fire was six inches long, the other four were twelve inches. A thermometer was pla.ced in each compartment, and one in the neck of the chimney, to indicate the tempera-t ture of the water in each chamber, and that * of the escaping products of combustion. In f) an experiment of four hours' duration, 6 S pounds of water were evaporated in the first compartment of six inches, 2 pounds 9 ounces in the second, 1J pounds in the third, 1 pound 3 ounces in the fourth, and 1 pound in the last. The small chamber nearest the fire evaporated five times more water than the one of double the size, at a distance of four feet further from the fire. The heat of the escaping products of combustion in the chimney was about 500 Fah. Another experiment was tried with the same boiler lengthened half a foot in order to make the water chamber nearest the fire of equal size with the other four. In an experiment of three hours' duration, 7 pounds 5 ounces of water were evaporated in the first compartment ; in the last, near the chimney, 3 pounds 15 ounces. This gives a far greater useful effect, for the water chamber farthest from the fire than in the first experiment. In the first, the compartment nearest the fire was ten times more efficient, in the second, not quite double. What was the reason of this ? It is stated that in the second experiment the heat of the escaping products of combustion was 800 Fah., or 300 higher than in the first experiment. It therefore appears, from these results, that in order to obtain anything like a useful effect from the heating surfaces of long boilers, that the products of combustion must escape at a very high temperature. " These experiments," says Mr. Williams, " dispel the illusion that mere tubular surface has the heat absorbing power usually attributed to it." Mr. Zerah Colburn and Mr. Thomas Presser have, in the American Railway Times, replied to Mr. Williams in reference to the efficiency of tubular boilers. To us it appears that the experiments referred to were only held by the author of them to demonstrate that mere heating snu-face, as commonly understood, is also improperly understood in regard to its evaporative efficiency. Thus, for example, one square foot of grate area, and nine square feet of heating surface is hel to be equal to a horse power. Regarding the amount of heating surface (that of a flue or tube in a boiler) allowed for a horse power, there aie a variety of opinions. Some hold that only one-third of the tube surface should be called effective ; others one-half (the upper), while others consider that the whole surface surrounded by water, or vice versa, should be calculated as efficient. Allowing any of these views to be held as a standard, the experiments referred to prove how incorrect such a standard is. The heating surface in boilers is efficient according to its proximity to the fire ; its efficiency decreases as it recedes from the fire, therefore it is not so much the amount as the position of the heating surfaces which should be considered efficient.