expansive direct acting steam pump requires, and the most of the steam is used simply to warm feed water.

To still further clear up the subject I will add, at the suggestion of Prof. Denton, who has kindly given me a number of useful hints as to the points to be made clear, some illustrative examples, bringing out the way in which the injector makes its wasting capacity evident, before which, however, there are two other points of Mr. Kent's criticism to reply to.

As to whether the injector raises water or takes it from a tank higher than the boiler does not affect the question in the least; it simply alters somewhat the amount of work to be done in transferring the water from the tank into the boiler, lifting the water even 15 or 20 feet is a small part of the work to be done in feeding a high pressure boiler, and the question under discussion is not just how much mechanical work there is to be done, but whether the injector can do it economically, and I think I have shown that it cannot.

As regards Mr. Kent's comparison of the injector with a steam trap, so far as the two instruments resemble each other, the criticism of the injector applies also to the trap. We may take some future occasion to discuss the exact action of a trap, and are having one set up at Stevens Institute for experimental purposes.

ILLUSTRATIVE EXAMPLES.

I. Suppose a boiler, capable of furnishing a maximum of 16 pounds of steam per minute, at 80 pounds pressure above the atmosphere when fed with water of 150° temperature, to be running an engine, out of which we desire to get as large a H. P. as possible to run a mill. A test is made of this H. P., and to make it as large as possible we feed the boiler by a hand pump, which forces the water through a heater in which it is heated from 60° to 150° by the exhaust from the engine. The work of feeding the boiler with 16 pounds of water is

and this being done by manual labor on the force pump is transmitted by a direct push through the mass of water and steam to the piston of the engine, where it reappears as part of the indicated

H. P.

Assume that the test shows that the utmost that can be done with boiler and engine is 32 H. P., all of which is now available to run the machinery of the mill.

II. Suppose now that the force pump is to be driven, as a regular thing, by a belt from the engine, then we shall have left for running the mill only 32.056 31.944 H. P.

=

III. Suppose that we provide in addition a non-expansive direct acting steam pump, which can be used when the geared pump is out of order and when the engine is not running. Such a pump is known to require about one-fourth of a pound of steam for every 16 pounds of water forced into an 80 pounds pressure boiler, using therefore one sixty-fourth of the whole supply of steam, and consequently reducing the performance of the engine by one sixtyfourth, so that we shall have for running the mill only 32 - .5 31.5 H. P.

=

IV. But the steam pump may fail and so an injector is put in. An injector takes about 1 pound of steam for every 16 pounds feed water drawn from the 60° supply and furnished to the boiler at 150°, and so our available steam is cut down from 16 pounds to 15 pounds per minute, and the H. P. of the engine from 32 to 30 H. P. To make sure that this is a correct calculation, we may remember that the utmost capacity of the boiler is to evaporate 16 pounds per minute from feed water at 150° into steam at 80 pounds, and that the injector does no more than furnish 16 pounds feed per minute at 150°, while it takes 1 pound of steam per minute to run it.

The heater is thrown out of use when the injector runs, and consequently the heat of exhaust otherwise used to warm the feed water, is allowed to waste. This waste equals 16 pounds × (150 - 60)° = 1440 B. T. U. per minute to be charged against the injector.

To see that the injector wastes temperature we observe in this case that the injector warms the water with live steam at 80 pounds while the heater does it with steam at say 18 pounds, now the exhaust steam is of no use to run the engine, because its temperature is too low, while the value of the 80 pounds steam in running the engine is in consequence of its high temperature and consequent high pressure, so that the temperature makes the difference between the valuable and the almost worthless steam and, therefore, in using the live steam, the injector wastes the higher temperature of that steam.

CCCXXVI.

ON THE IDENTIFICATION OF DRY STEAM.

BY JAMES E. DENTON, HOBOKEN, N. J.

(Member of the Society.)

INTRODUCTION.

DRY steam is understood to be saturated* steam corresponding to a given pressure, and the latter is understood to be identified by the relation between pressure and temperature and latent heat, determined by Regnault's experiments, the results of which are presented in tabular form in all publications upon the properties of steam. For example, saturated* steam for 90 pounds pressure per square inch should be at 320 degrees temperature, Fahr., and should possess latent heat equal to 808 British thermal units. We also know, through the labors of Messrs. Fairbairn & Tate, that such

* The term "saturated," as applied to steam, appears to be sometimes understood as referring to a condition of wetness, whereas it implies the most perfectly gaseous condition of steam possible without the existence of superheating. The term "saturated" originates in the presentation of the laws of vapors in treatises on physics, where the vapors of water, ether, etc., are supposed to be confined in a space above a surface of some liquid, such as mercury, other than that belonging to the vapors. If water is introduced, drop by drop, into a space at the top of a closed mercury column, at say 60° Fahr. where less than the pressure of the atmosphere prevails, such water will flash into vapor until the space is under a tension equal to the pressure of steam corresponding to 60° temperature. Then if more water be introduced into the space, it refuses to vaporize, but accumulates as liquid water on the surface of the mercury, and consequently the space, and hence the vapor in that space, is said to be "saturated." Before the space or vapor is thus saturated, the vapor of water present is "non-saturated" steam, and if compressed, its pressure increases without causing any liquefaction, the vapor following the laws of fixed gases, like air, etc. When the space or vapor becomes saturated, any compression of the vapor does not result in increased pressure (the temperature being assumed constant), but instead some vapor liquefies. Similarly the steam in a practical boiler (where there is always liquid water beneath the steam) is saturated, because any effort to make a given weight of steam occupy less space, either by raising the water level or by other compression of the steam, causes a portion of this weight of steam to liquefy without changing the vapor tension, assuming the temperature of the contents of the boiler to remain constant. The only condition at all practical corresponding to "non-saturation" as described in physics is when steam is superheated.

steam weighs 0.207 pounds per cubic foot, or that this figure is its density in pounds. If a boiler is steadily generating and delivering to an engine steam possessing exactly these qualities and the water under the steam be violently disturbed, its liquid particles may mingle with the gaseous particles of the steam, and a pound of the mixture formed will no longer possess the same latent heat or density, yet the pressure and temperature will still be the same as that of the exactly saturated steam. Such steam is practically known as "wet" steam, and in contradistinction the term "dry steam" has arisen, the latter meaning simply exactly saturated steam. If a sufficient portion of the heating surface of the boiler above the water-line be exposed to the action of the fire, the pressure of the steam may remain the same, and yet its temperature may be greater, the latent heat greater, and the density less than corresponds to saturated steam. Such steam is practically known as superheated

steam.

In measuring the performance of a boiler, the essential determination is the quantity of heat utilized by the generation of steam. If the steam generated at say 90 pounds pressure is dry steam, then for each pound of feed water the boiler is to be credited with utilizing 120 heat units, due to the temperature of the steam if the feed water is at 200° Fahr., and 808 heat units due to its latent heat, or a total of 928 heat units. If, however, 10 per cent. of the steam is liquid water mechanically mixed with 90 per cent. of dry steam, then for each pound of feed water the boiler is to be credited with 1.10 × 120 heat units, due to temperature, and 0.90 x 808 heat units, due to latent heat, or a total of 859 heat units, which is 92 per cent. of the dry steam total. Unless, therefore, allowance for the presence of moisture is made, the efficiency of a boiler is made too great for ordinary steam pressures, at the rate of per cent. for each one per cent. of water in the steam. Again, if steam at 90 pounds pressure is superheated 10° Fahrenheit, so that its temperature is 330° F., then for each pound of feed water at 200° F. we must credit the boiler with the heat due to dry steam plus 0.48 × 10° 4.8 heat units, so that failure to allow for superheating makes the efficiency of a boiler, at ordinary pressures, too low by about 0.05 per cent. for each degree Fahrenheit of superheating.

It is customary among experts to make these allowances in reporting the performances of boilers, and hence arises the necessity of determining to what extent the steam generated by a given boiler differs from exactly dry steam.

If the steam is superheated, the simple observance of its temperature by a proper thermometer affords the desired data. If, however, the steam is shown by a thermometer to be at exactly the temperature due to saturation, it may contain any amount of water in suspension, and the determination of the amount of the latter can in general only be accurately known by a measurement of either the latent heat or density of a known weight of the mixture, the determination of the density is an operation too delicate to have been yet attempted with portable apparatus. The determination of latent heat involves simply the condensation or mixture of a known weight of steam in or with a known weight of some other substance of known specific heat, and the operations to be performed are such as can be carried out with apparatus of a conveniently portable nature. Nevertheless attempts to use portable apparatus or calorimeters for the determination of the latent heat of steam have, in the main, been very unsatisfactory, and opinions are divided among experts whether it is best to seek other methods than that of the condensing calorimeter or to attribute the latter's unsatisfactory results to unskillful use.

The object of the following investigations is to contribute material to both sides of this question.

1st. By proposing a method, based upon experiment, of recognizing dry, slightly wet, or slightly superheated steam by the scrutiny of a jet of steam flowing into the atmosphere.

2d. By a theoretical discussion of the instrumental errors to which condensing calorimeters are liable.

PART I.

EXPERIMENTS WITH STEAM JETS.

If a boiler can be made to generate steam which is a few degrees superheated, then by drawing off steam at the end of a pipe of sufficient length the loss of heat by the pipe may be made to so nearly equal the amount of the superheating that the steam will issue from the pipe in exactly the saturated condition. In the case of these experiments, this method was adopted to obtain dry

steam.

A 30 HP. Harrison steam boiler, Fig. 59, was used, which, when not forced to its utmost steaming capacity, superheated its steam from six to twelve degrees Fahrenheit. To the top of the steam