This water was alkaline, but its alkalinity, which was due to the magnesia and, probably, lime resulting from the decomposition of the carbonates of these two bases, was lost by the calcination of the solid matters dissolved in it. These two oxides are found in the solid residue in presence of organic matter which during the circulation furnished them with carbonic acid.

An examination of the water of condensation drawn from the surface condenser showed it to be strongly acid and containing iron, while the water in the boiler from which it was distilled was alkaline and free of iron. The acid was found to be carbonic acid; the iron was in the state of carbonate of iron, and came from the iron tubes of the surface condenser. The action on these tubes was so rapid that the clear water of condensation after standing one night in the condenser was discolored and ochrous in the morning. This water, after the removal of the acid and iron, was absolutely pure. If the water was allowed to stand, the acid left it spontaneously, and at the end of about fifteen days was nearly all gone.

The writer has observed the same destructive action of the water of condensation on the cast iron shell of the surface condensers of marine engines, due to the gaseous acids brought over with the steam from the boiler and dissolved anew in that water. It is quite an error to suppose the water of condensation obtained from a surface condenser in ordinary use is pure distilled water: it is both acid and ochrous.

That no water was "primed" over, or entrained by the steam during the experiments, was due to the large capacity of the steam-drums and to their height above the water level, as other boilers of the same type, not so well provided in these particulars, are found to prime considerably, though such steam-drums do not belong to the theory of this kind of boilers.

One of the great advantages of a boiler like the one experimented

with, in which the water is contained in small tubes, is its less probability of explosion, and the less destructive effects should an explosion happen, than in the case of boilers with either cylindrical shells or rectangular shells stayed at intervals. The less probability of explosion results from the fact that the small tube is stronger to resist a given internal pressure than a large cylinder or stayed flat surface. The less destructive effects of an explosion, should one occur, results from the fact that the boiler composed of small tubes contains less water for a given heating surface than those whose shells are cylindrical or rectangular.

The explosion of a boiler is not instantaneous; it is progressive. Its more or less violence depends on the force accumulated in the boiler, and on the duration of the explosion itself. The force accumulated depends almost exclusively on the quantity and temperature of the water. Suppose, for example, that a cubic foot of water heated in a boiler to the temperature corresponding with a pressure of 60 pounds per square inch above the atmosphere, should be suddenly discharged into the air, it would produce about 185 cubic feet of steam of atmospheric pressure. Let the volume of steam in the steam-room of a boiler under the supposed pressure be the greatest given in practice, it will bear but a feeble proportion when expanded to the atmospheric pressure to the volume which will be produced from the water in that boiler, and it will contribute to the destructive effects of the explosion only in that proportion. Practically, the distance to which the destructive effects of a boiler explosion extends, and the intensity of the destruction, increase in proportion to the quantity of water contained in the boiler and to the temperature of that water.

When one or more plates of a boiler are ruptured by the pressure, a considerable fraction of its contained water passes into steam under the lessened pressure thus produced, and this transformation takes place suddenly on all the particles of the water, in measure and to the extent as they emerge into the gradually lessening pressure; so that it depends on the size and situation of the first rupture relatively to the objects in front, whether the steam they thus encounter is sufficient to destroy them. The explosion of a boiler containing but little water may sometimes kill those who are in close proximity, but the destructive effects are small and limited to the near locality. When, on the contrary, the quantity of water in the boiler is large, the surrounding WHOLE NO. VOL. CX.—(THIRD SERIES, Vol. lxxx.)

27

objects and the fragments of the boiler itself are violently hurled by the explosion and its destructive effects extended into remote places.

From the foregoing considerations it would seem that, in function of safety, boilers should be composed of separate similar parts, the more numerous the better, other things equal, and that each part should contain the minimum of water, conditions which are well fulfilled in the boiler experimented with, whose every vertical row of connected tubes may be considered as a complete and separate system; so that in the event of an explosion there would be but little water to cause destruction, and that the water in the different parts would flow slowly and in succession towards the fracture for projection into the air, thus rendering the discharge as prolonged as possible, and the effect, consequently, a minimum in intensity.

These conditions, unfortunately, have but a limited application. The free circulation of the water in the restricted separate parts is much impeded, and the small quantity of water itself becomes a serious defect in view of the difficulty of maintaining with it a uniform pressure of steam. The less the mass of water the less constant is the pressure.

When the fires are urged beyond a certain point the steam is so rapidly formed in the tubes, proportionally to the vent for its escape, that notwithstanding the considerable inclination given to them, the water is driven out at both ends, to return again, intermittingly, upon the metal overheated in its intervals of absence. As the tubes lie directly over the grate, the heat comes equally upon nearly the whole length of the four lower rows, so that there is nothing but their inclination to produce a circulation of water. These lower rows are impinged on by the gases of combustion at their maximum temperature as they first rise from the incandescent coal whose radiant heats also acts at the same time equally upon their whole length, so that the generation of steam must be enormously rapid there, while the only means of producing a circulation or water current through these tubes is the difference in the heads of water at their two ends.

This slight provision for establishing a circulation is neutralized in sea-going steamers by their pitching, rolling or permanent heeling over under sail, so that this type of boiler is absolutely precluded from marine use on that account. Under favorable conditions it can be employed for land service, but if attempted to be used at sea will speedily be itself destroyed, and probably cause grave accidents.

The slipping on each other of the cast iron boxes at the front and back of the boiler, into which the tubes are secured by pairs, makes a provision for any unequal expansion and contraction they may undergo from differences of temperature. But, as these free joints to move. are also free to admit the cold atmospheric air, the entire front and back of the boiler should be enclosed by air-tight doors, which, though adding to the cost of constructing the boiler, would materially lessen the weight of coal consumed to produce a given quantity of steam in a given time.

The heating surface per unit of area in boilers of this description, as compared with that of boilers of the usual construction, is very inefficient. Less water, other things equal, will be vaporized in them per square foot of surface, because all the stem generated in the lower tubes has to force its way, successively, through the upper tubes, alternately, from end to end, displacing to that extent the water in the latter and thus preventing its contact with the metal. In other words, less water is in contact with each unit of area of heating surface in boilers of this kind than in boilers of the ordinary construction. Hence, to obtain an equal economic vaporization, a much greater proportion of heating surface to grate surface must be given, or a much less rate of combustion adopted. In the experimental boiler there were 49.1153 square feet of heating surface to one square foot of grate surface, a proportion approaching what is given in many locomotives with their enormous rate of combustion. A well designed ordinary boiler with half this proportion of heating to grate surface, and the experimental rate of combustion, would have given equal economic and potential results. Real improvement in engineering consists in accomplishing equal effects with less expenditure of material and labor.

Digestive Ferments in Vegetables.-M. Bouchard was led to believe, by his discovery of papaine, the digestive ferment of the Carica papaya, that there might be a general carnivorous property in the latex of many other vegetables. Some special careful investigations confirmed him in this belief, and his experiments with the milky juice of the common fig tree showed that it contained powerful digestive properties. He proposes to continue his researches in order to discover the nature of the new principle of vegetable pepsin.-Comptes Rendus.

C.

THE WEAKENING OF STEAM BOILERS BY CUTTING HOLES IN THE SHELL FOR DOMES AND NECKS.

By W. BARNET LE VAN.
(Continued from page 323.)

The association above referred to at the time the "French" boilers were offered for their guarantee, had under consideration several instances of explosion, which had been attributed by their chief engineer to the employment of such necks, and to large openings not sufficiently strengthened, led the committee to the determination to try the matter fully. Indeed, so completely were they alive to the weakness resulting from the large holes at the base of steam domes, and the presence of domes at all, that they had for years urged members when laying down new boilers, to dispense with domes altogether.

It was, therefore, arranged that a proper boiler should be constructed, of the diameter adopted in daily practice, and of the usual thickness of plates, with actual manhole, mouthpieces, etc., such as are common, so that the ultimate test should be decisive.

With this object then in view, they had a boiler constructed 21 feet long by 7 feet in diameter inside the inner plate of shell; with two furnace tubes 2 feet 9 inches inside diameter with flanged seams, each ring being welded up solid so that there would be no rivets or lap joint in the flue. The shell plates were inch thick, the ends