formation of CO results in a very serious loss of heat, and must therefore be prevented by admitting some excess of air.

The excess of oxygen required is generally from 6 to 8 per cent. of the volume of the gases. If there is less than 6 per cent. of oxygen there will almost certainly be traces of CO.

The Orsat apparatus enables the percentages of oxygen, carbon dioxide, and carbonic oxide to be ascertained directly. The remainder is usually considered to be nitrogen, as, although there are traces of other gases, they are insignificant.

The apparatus, which is shown on page 73, consists essentially of a measuring tube A, into which a sample of the gas is drawn, and of three other vessels B, B1, and B2, which contain substances capable of absorbing respectively, carbon dioxide, oxygen, and carbonic oxide.

The method of using the apparatus is as follows:

Through a suitable hole in the chimney, uptake, or flue, insert a piece of iron tube, long enough to reach well past the center, the tube having saw slits in its circumferential plane for a length of 12 inches or more. If desired, a tube perforated with small holes may be used instead.

See that the aperture in the chimney, round the tube, is tightly plugged, so as to prevent air (which would probably vitiate the results obtained) being drawn in.

Place the apparatus in a convenient position near the chimney, the bottom of the apparatus being, say, about 3 feet above the level of the feet of the observer; connect the end of the iron tube to the apparatus by an india-rubber pipe D, having a U-tube filled with glass wool inserted at the position marked E, between the apparatus and the boiler, so as to intercept dust.

The bottle C is to be filled about two-thirds full of water, and connected to the bottom of the measuring tube A by an india-rubber tube. When this bottle is placed on the top of the case containing the apparatus, or at some other convenient similar height, the water will naturally flow into the vessel A.

If now the bottle C be placed below the apparatus and the cock a opened, it is evident that as the water flows out of A the gas will be drawn in from the flue and take its place. Draw in the gas well below the zero mark, and cut off the connection with the flue by closing the cock a. Then lift the bottle C, so that the water level in it coincides with the zero mark in the measuring tube, and open the three-way cock a to the atmosphere to allow of the surplus gas escaping. We thus obtain the tube A full of gas at atmospheric pressure. Again close the cock a. Then, by opening one of the cocks, b, b1, or b2, the gas contained in the measuring tube A can be forced into either of the vessels B, B1, or B2, by raising the bottle C so that the water flows into A, due care being taken that the water never

rises above the mark at the top of the measuring tube. The vessels B, BI, and B2 contain the following reagents:

These absorbing vessels should be filled with the reagents, rather more than half-way up.

It is essential that the gases to be tested be passed through the different reagents in the order given above, otherwise incorrect results will be obtained.

The vessels B, B1, and B2 contain small glass tubes. These are used with the object of giving a greater wetted surface to absorb the gas introduced. The tubes with copper wire round them are for the vessel B2 containing cuprous chloride.

NOTE. Care should be taken to keep the pyrogallic solution from air, as it absorbs oxygen rapidly. It is best to mix the potash solution with it in the tube.

The measuring tube A is, for convenience of calculation, marked off into 100 parts, so that percentages may be read off easily.

At the moment of measuring the volume of gas in the graduated tube, the water bottle must be held at such a height that the level of the water in it is exactly the same as in the graduated tube, otherwise the gas will be compressed or expanded by the difference between the two columns of water.

Before commencing the test get rid of as much as possible of the air in the tubes by using the small hand-bellows shown in figure; then draw several samples of the gas into the measuring tube, and discharge each in its turn to the atmosphere through the three-way cock a. Having obtained an undiluted sample, shut the cock a, open the cock b, and force the gas into the vessel B. Draw the gas back into the vessel A, and repeat the operation three or four times, so as to ensure the thorough absorption of the CO2. The last two readings should give the same result, showing that the absorption is complete.

Follow the same procedure with the remaining two vessels B1 and B2, taking the reading of the reduced quantity of gas in the vessel A after each operation.

2

CO, is absorbed by the caustic potash solution very quickly, and it will be found that passing the gas three times through the absorbing vessel B will generally be quite sufficient. The gas, however, must be passed through the pyrogallic solution at least five or six times, in order that

the oxygen may be all absorbed. If this be not done, the oxygen remaining will be absorbed by the cuprous chloride, and will be mistaken for CO, although there may be none of that gas present.

The total of the percentages of the three gases CO2, CO, and O, should be about 19.5, and this rule may be used as a rough check on the analysis. As the percentage by volume of oxygen in air is 21, the volume of air corresponding with any given volume of oxygen may be found by multiplying by 100, or 4.762. The volume of air corresponding to a given volume of CO, may also be found by multiplying by the same figures.

EXAMPLE:-Analysis shows..

13.5 CO2
6%
O

Then air used for combustion =
And excess air

13.5 × 4.762 = 64.3
= 6 × 4.762 = 28.6
92.9

The percentage of excess air above that which is necessary for combustion is therefore:

Care should be taken with regard to the following points:

I. The absorbent should not be forced below the point D, or some of the gas may escape and be lost, and, of course, an incorrect result obtained. 2. The absorbent must be at exactly the same level in the tube--say at C, when measuring the volume after the gas has been absorbed as before. 3. Time must be allowed for the water to drain down the sides of the tube before taking a reading. The time must be the same on each occasion, otherwise more water will drain down at one time than another, and an incorrect reading result.

REASONS FOR HIGH EFFICIENCY OF BOILERS

T

HE text-books on physics explain that fluids are heated (and cooled) not by direct conduction of heat, but by what is called convection, where the portion near the source of heat has its temperature raised and is displaced by the cooler portions which are heated in turn. It is evident, therefore (as already referred to on page 29), that the boiler which provides most thoroughly definite paths for the hot gases and the water and, at the same time, breaks them up so that all portions can intermingle, will abstract the greatest percentage of heat and thus give the highest efficiency. The Babcock & Wilcox boiler accomplishes this in the most thorough manner.

By means of the fire-brick roof and the increasing volume of the furnace, the fuel is given ample opportunity for complete combustion before the hot gases pass among the tubes. By means of the vertical baffles, the gases are directed at right angles across the tubes three times; and, in addition, owing to the sinuous headers which "stagger" the tubes, the gases are thoroughly broken up and every part brought in contact with the heating surface. From the top of the last pass, the gases go direct to the smoke pipe, with a minimum loss of draft. It will be seen, therefore, that the circulation of the gases is perfect.

The water is fed into the steam and water drum and thence descends through the connecting nipples to the front headers, from which it passes through the tubes receiving heat and being partly converted into steam. The mixed steam and water fills the back headers and passes through the return circulating tubes to the steam and water drum, separating so that the water falls into the body of water in the drum, while the steam passes around the ends of the baffle-plate into the steam space. It is hard to conceive of a simpler and more direct circulation.

Inasmuch as boilers have been constructed with tubes at every inclination from horizontal to vertical, the query would naturally arise whether one inclination is better than another. This subject has been investigated by special tests and confirmed in practice, showing that maximum results are secured when the tubes have an inclination of 10 degrees to 15 degrees to the horizontal. The standard angle for the Babcock & Wilcox boiler is 15 degrees.

The increased rates of combustion with coal and the use of oil fuel, with its possibilities of very high forcing, have raised the question of the effect of this greatly stimulated evaporation on the circulation, that is, whether it will be as definite as at lower rates but simply increased in amount. This has been the subject of careful laboratory investigation and also of extended tests of boilers under all degrees of forcing. These have all shown conclusively that, in a well-designed boiler with a simple