water, inside of twelve years, and frequent repairs will be necessary at the joints and connections.

We speak of the sectional water tube boiler as it is generally designed and constructed. Further on we shall speak of the sectional system at length, though we will here say that we believe that no water tube should be less than 5′′ in diameter; else, when arranged on the sectional plan, the water will be driven from the lower courses of tubes and the metal burned to an injurious degree. When a shell fire tube boiler is well designed and proportioned, nine (9) square feet of heating surface will produce a horse power, with an allowance of one (1) square foot of grate surface for every thirty-two (32) square feet of heating surface.

FIXING THE NUMBER OF TUBES.

We have already stated that most of the boilers now in operation have too many tubes for the best economy.

This conclusion was reached by Mr. Hoffman by actual experiment. Of fifteen boilers experimented on, all using well proportioned heaters, and above the average steaming coal, only two gave an evaporation of nine pounds of water per pound of fuel, and only one 9ths pounds. Four touched 81⁄2 pounds, two 8%, and the remaining six ranged from 8 down to 7%, which was the lowest.

The highest evaporation was reached with a 66" boiler, containing sixty-six 31⁄2" tubes, fourteen feet long; eight of the lower courses of tubes, and four of the side courses being purposely closed at each end. Before the change was made, this boiler gave an evaporation of 8% pounds to one of the fuel, and after the change the evaporation was 9 pounds.

Before the change the tube area was grate surface, and after the change area compared with grate.

of the was the

The heating surface before the change was 39 to one, compared with grate, and after the change, 354 to one. The temperature at the flue leading to chimney (chimney 120 feet high) before the change was 462°; by the change this was raised to 540°, and the damper was shifted from three-fourths open to half closed. The velocity of the hot air was, before the change, about 122 feet; after the change, this was increased to 136 feet per second. This goes to

show that before the change there was too much tube area for the grate (although the grate surface was a little too small). The extra amount of tube surface offering too much resistance, by friction caused the currents of hot air to be sluggish, and, of course, the temperature fell below the economical point, and the heat did not impinge on the sides of the tubes with a sufficient degree of intensity. We think it is generally

conceded that the gases cannot leave the boiler

Draft in

at a point below 520°, without impairing the f

draught power of the chimney.

In the case under notice, this power measured in inches of water 1.112; this was after the number of tubes had been reduced. The cold air entered at about 70°.

Therefore, we believe that the tube area should be about of the grate area, and sub-divided as } little as possible, consistent with good circulation, and making no tube more than 48 diameters long.

The boiler giving the lowest evaporation had a 42-inch shell, and contained seventy 21⁄2 V tubes, staggered tubes being fourteen feet long. This boiler was primed continually, and none of the usual remedies seemed to affect it. The tubes were 3/4 of an inch apart, and, under ordinary firing, a vibration of the tubes could be noticed, caused, of course, by the forced currents of water through cramped passages.

Other boilers with sixty-four 31⁄2 inch tubes 18 feet long, gave only an evaporation of 8% pounds; and of those with 3′′ tubes, 16 feet long (60" shell), 74 tubes gave 8 pounds, and one of the same size with tubes, staggered or zig-zag, 7 pounds.

The fuel in all cases was hard coal of various brands, of quite uniform quality, Buck Mountain Lehigh being used in the case of 60" shell, 74 tubes, 16 feet long, and the 66" shell, 54 tubes, 14

area.

feet long. The number of tubes for a certain diameter of shell is, therefore, limited, for several reasons. Among these is the matter of greate A line representing a point equidistant between and the total area of grate has proved to be the most efficient, as connected with the ordinary factory chimney. Then there is the matter of uniform and free circulation of the water in the boiler, and having too many tubes prevents the much desired result.

The gases must have a free, and, to a certain extent, rapid passage through the tubes, and must not flow at a velocity less than 132 feet per second, where they leave the boiler.

Then another item that has been overlooked, is the amount of water capacity at the points where the heat is the most intense.

Boilers with tubes too near the bottom and sides of shell repel the water, instead of converting it into steam, and in so doing retard the circulation almost entirely at the surfaces, where most of the steam is or should be generated; hence, the low evaporative power of this style.

Besides, the trouble with too many tubes, is the great trouble of cleaning; having poor circulation, the deposits collect with alarming rapidity, and in a short time the boiler is as good as ruined; and it cannot be reclaimed, as it is impossible to clean it.

There is just one remedy, and that is to remove some of the tubes.

In support of this theory regarding large tubes, and a limited number of them, we will call to the minds of engineers who practised twenty years ago, the stationary return drop flue boilers, used so extensively at that date in the Eastern States. A 66" shell boiler would take the gases direct from the furnace through four flues, and return them to the front connection through two flues. Their average evaporation was 9 to 1; but they were expensive to build, and rather weak in construction at some points.

SHELLS AND TUBES

The shells for all fire tube boilers should be as large as consistent with safety, at the ordinary working pressures, in order to expose as large an area of water level as possible, as this will help to prevent priming.

In regard to heating and grate surface, our proportion is from 32 to 34 square feet of heating surface to one (1) square foot of grate surface or

area.

Under no circumstances more than 36 square feet of heating surface to one (1) square foot of grate with natural draught.

The combined tube opening should be no more than one-fifth the total grate area, and not less than one-eighth of the same. The tube opening should never with natural draught be subdivided more than sixty times, and fifty-four tubes will generally give the best results.