Scientific American Supplement, No. 799, April 25, 1891

Part 1 out of 2

Produced by by Don Kretz, Juliet Sutherland, Charles Franks and the
DP Team



NEW YORK, APRIL 25, 1891

Scientific American Supplement. Vol. XXXI, No. 799.

Scientific American established 1845

Scientific American Supplement, $5 a year.

Scientific American and Supplement, $7 a year.

* * * * *


I. ARCHITECTURE.--Marble and Mosaic.--By T.R. SPENCE.--A
paper recently read before the Architectural Association, London,
containing valuable suggestions for designers of buildings.

The St. Lawrence Hospital for the Insane.--A New York State
hospital recently built from designs by the State architect.--Full
description.--1 illustration.

II. BOTANY.--Lavender and its Varieties.--The history, properties,
and technology of this plant.--2 illustrations.

III. CHEMISTRY.--A Projecting Apparatus of Precision.--A useful
adjunct for the chemist's balance, accelerating the operation of
weighing.--1 illustration.

Spectrum of the Sun and Elements.

Allotropic Forms of Metals.

IV. HYDRAULICS.--The Power of Water, or Hydraulics Simplified.
--By G.D. HISCOX.--Current wheels for power and raising water.
Interesting presentation of this practical portion of the subject.
--4 illustrations.

V. MECHANICAL ENGINEERING.--Compressed Air Production.--
By WM. L. SAUNDERS.--A Sibley College lecture, giving full
elaboration to this important subject.--The various forms of
compressors and general features of the service.--18 illustrations.

Improved Pneumatic Hammer.--A suspension hammer capable
of delivering 500 blows per minute.--8 illustrations.

The Thermic Motor of the Future?--A remarkable exposition of
the possibilities of the gas engine.--Recent experiments under M.
Aime Witz.--2 illustrations.

VI. MEDICINE AND HYGIENE.--The Electrical Purification of
Sewage and Contaminated Water.--By WM. WEBSTER.

VII. NAVAL ENGINEERING.--The New German Dispatch Boat
Meteor.--A German built war vessel of great speed.--Her dimensions
and rating.--1 illustration.

The Raising of the Ulunda.--A remarkable feat.--The raising of
a steamship sunk off Nova Scotia.--10 illustrations.

VIII. TECHNOLOGY.--Starches for the Finishing of Cotton Fabrics.
--Classification of starches, with illustrations of their appearance
under the microscope.


In time of war the dispatch boats are the eyes of the fleet. It is their
duty to reconnoiter and ascertain the strength of the enemy and to carry
the orders of the commander. For this service great speed is of the
utmost importance. As all nations have increased the speed of their war
ships during the last few years, it has become necessary to build faster
dispatch boats. Although our new vessels of this class, Blitz, Pfeil,
Greif, Jagd, and Wacht, fulfill the requirements, still greater speed
was deemed requisite, and steps were taken for the construction of the
Meteor, which was launched at Kiel in 1890. This vessel is 262 ft. long,
31 ft. wide, and has a draught of 13 ft., and a displacement of 950
tons. There are two independent engines, each of which develops 2,500
h.p., making a total of 5,000 h.p.; and each engine drives a screw. When
both engines are running with their full power, the Meteor has a speed
of 24 knots (over 271/2 miles) an hour, which is equal to the speed of a
freight train.[1] As the resistance of the water increases greatly with
an increase in the speed of the vessel, the engines of the Meteor are
very large in comparison with the size of the vessel. The largest
armored vessel in the navy, the Konig Wilhelm, for example, has a
displacement of 9,557 tons, and its engines develop 8,000 h.p., driving
the vessel at a rate of 14 knots an hour; that is, 0.84 h.p. to each ton
of displacement, while in the Meteor there is 5.26 h.p. to each ton of
displacement. The Meteor has a crew of 90 men, and an armament of eight
light guns, and has no rigging; only one mast for signaling. Steam power
is used for raising the anchor, removing the ashes from the engine room,
and for distilling water. The vessel is lighted with electricity, and is
also provided with electrical apparatus for search lights.--_Illustrirte

[Footnote 1: This, we believe, is the fastest vessel of the kind
afloat.--ED. S.A.]


* * * * *


Shortly after the recovery of the Ulunda, below described, the North
American and West Indian squadron of the Royal Navy visited Halifax,
Nova Scotia. The simple and novel means adopted for raising the ship
attracted considerable attention among the officers of the fleet, and by
way of stimulating the studies of the junior officers in this branch of
their duties, a prize was offered for the best essay on the subject, to
be competed for by the midshipmen of the various ships. The essays were
adjudicated upon by Captain W.G. Stopford, of the flag ship--H.M.S.
Bellerophon--and the first prize was awarded to the following paper,
written by Mr. A. Gordon Smith, of H.M.S. Canada. The article needs no
apology, but it is only just, says the _Engineer_, to mention the fact
that the writer is not yet eighteen years of age.

The steamship Ulunda, on the remarkable raising and recovery of which
this paper is written, is an iron screw ship of 1,161 tons, until lately
belonging to the Furness line. It is a sister ship to the Damara, of the
same company, and was built and engined by Alex. Stephens, shipbuilder
and engineer, at Glasgow, being fitted with compound vertical engines,
of 200 nominal horse power, having two cylinders of 33 inches and 66
inches diameter respectively, which are capable of sixty-five
revolutions per minute, and give a speed of twelve knots an hour.

For supplying steam to the engines there are two return-tube boilers,
each having three furnaces, and there is also a donkey boiler, which is
used in harbor for working the four steam winches on deck.

She is divided into seven watertight compartments by athwartship
bulkheads. The foremost one is the usual collision bulkhead. Between
this and the foremost engine room bulkhead are Nos. 1 and 2 holds,
separated by a watertight bulkhead. Abaft the after engine room are two
more holds, divided in the same manner as the forward ones, and astern
is another compartment, in which all stores are kept. Coal bunkers form
a protection for the engines and boilers. Fore and aft the ship, as low
down as possible, are a number of ballast tanks, which can be filled
with or emptied of water as occasion requires to alter the trim of the
ship. Extending over all holds there is a strong iron lower deck, about
8 feet below the upper deck, which is pierced with a hatch over each
hold immediately under a corresponding hatch in the upper deck, for
stowing and unstowing cargo.


In the engine room there are six steam pumps, two of them bilge pumps,
worked by the main crossheads, for clearing the engine room of water.
For pumping out the ballast tanks there are two more, which have their
own independent engines. The remaining two are for various purposes.
Besides these there are several hand pumps on the upper deck.

Having been built in 1885, the Ulunda is almost a new ship, and has been
used principally as a cargo steamer, though she is provided also with a
saloon and staterooms for a few passengers. She was on her way from St.
John, New Brunswick, to Halifax, when during a thick fog she struck on
Cowl Ledge, a reef between Bryer and Long Islands, on the southwest
coast of Nova Scotia, about half a mile from the shore. The cause of the
disaster was probably one of the strong tide eddies which exist in the
Bay of Fundy, and which had set her in toward the shore. It was calm at
the time, and she was making seven knots an hour; and, being close to
the shore, leads should have been going in the chains. Had this
precaution been taken, very probably she would have been able to stop or
anchor in time to avert this catastrophe. There was no cargo on board,
it being intended to ship one at Halifax for London.

When ashore on this reef she was sold by public auction at Halifax, and
fell to a syndicate of private individuals for L440. These gentlemen at
once decided to raise her if possible, transport her into dock, and
repair her. They commissioned Captain Kelly, of the Princess Beatrice, a
ship then in harbor, to visit her and see what could be done for that
purpose. He went with a hired crew to Annapolis, and from thence
proceeded to the steamer by means of a tug, a distance of about forty
miles. When they arrived they found the Ulunda with her head to sea, and
her stern in only 2 ft. of water at low tide, with a list of 30 deg. to
port and her foremast broken short off. At high tide the water flowed
over the upper deck. On examination, the engine room was found full of
water, which did not rise and fall with the tide, showing that it had
been filled at high tide through its skylight. No. 3 hold was also full,
but had a slight leak, which was shown by the water falling slowly at
low tide and rising in the same manner at high water. The other three
holds were hopelessly leaky.

Upon investigation, it was decided to pump out the engine room
compartment and No. 3 hold, and to make the iron lower deck watertight
over the remaining holds. For this purpose three powerful pumps, with
the necessary boilers, were obtained from Halifax, sent by rail to
Annapolis, and then shipped on board a tug, from which they were hoisted
into the Ulunda by means of the derricks on the mainmast. These were
centrifugal pumps, capable of discharging 2,000 gallons a minute each.
One was placed in the engine room, another with its suction in No. 3
hold, and when these two compartments were pumped dry, it was found that
in No. 3 hold the leak was easily kept under, while in the engine room
there was no leak at all. The third pump was not used.


In the two foremost holds 2,000 large casks were then placed, and all
the hatches over the leaky holds--Nos. 1, 2, and 4--were battened down,
and made airtight with felt, pitch, tow, etc. A small hole was then made
in Nos. 1 and 2 hatches, about 2 ft. square. When the tide had sunk its
farthest, these two holes were closed and made perfectly airtight, in
the same manner as the hatches had been.

Before this took place the whole of the lower deck over the badly
damaged holds had been prevented from bursting up by means of wooden
shores, which were placed in rows about 4 ft. apart, and wedged firmly
into position. The wood for the shores was obtained from Annapolis, and
the casks from St. John. The ship went ashore on August 26, 1890. This
work was commenced on September 8, and completed ten days afterward.

The labor of repairing her could only be carried out at low tide, and
only then with the greatest difficulty, as the decks were very slippery
with weeds, etc., and inclined at an angle of 30 deg. Everything was
ready for floating her off at high tide on the 18th, and the hatches
were closed up on that day.

She was raised off the rocks by the water rising and compressing the air
in the two foremost holds, assisted by the buoyancy of the engine room
and No. 3 compartments. At high water the bow was afloat, but she was
aground by the stern. When, however, she was taken in tow by three tugs,
she slowly slid down the reef and floated into deep water. One tug was
placed on each bow, and the third was ahead. In this state she was towed
into West Port, a distance of four miles, and there beached on a
sheltered stretch of sand.

The casks performed no part in floating the ship off, but were only
there in case the great pressure of air should cause the escape of some
of it, in which event all the space underneath the lower deck would soon
have been occupied with water instead of air. These casks would then, of
course, have served to displace a large amount of this water, and so
keep her afloat. Luckily the deck did not leak, and the barrels were
thus not instrumental in the raising.

When beached the hatches were taken off, the casks removed, and a false
deck was built about 7 ft. below the lower deck, and about 10 ft. above
the keel. This was used as the bottom of the ship to take her round to
Halifax, and was built in the following manner: A kind of iron platform,
about 2 ft. wide, runs along the sides of the holds in the Ulunda for
strengthening purposes, braced at intervals of 15 ft. by iron beams
across the ship.

On this was built the wooden deck. Beams for this deck were constructed
of three 3 in. planks, and were laid down on the iron platform about 31/2
ft. apart, and firmly wedged into the ship's side. On these beams a
layer of 3 in. planks was placed in a fore-and-aft direction and nailed
down; on this were three layers of felt, and on this again more planks
were laid down in the same direction as before.

The whole deck was then carefully calked and the sides made watertight
with Portland cement. This deck only extended to the engine room
bulkhead through the two foremost holds. It was prevented from bursting
up by the pressure on the bottom of it, by means of shores, in the same
manner as the iron deck had been served before. Shores were, therefore,
connecting the three decks--the upper deck, lower deck, and wooden
deck--this being done to equalize the pressure on the _extempore_ deck
and the two permanent decks, and thus gain additional strength.

No deck was built in either of the after compartments, inasmuch as No. 3
hold was kept clear of water as before by its pump, and in No. 4 the
deck was not necessary. To have built one there, as in the two foremost
ones, although it would have given a little more reserve of buoyancy to
the ship, would have raised the stern higher than the bows, and so would
have increased the upward pressure on the wooden deck, and thus have
increased the liability to burst up. For the same reason, when raising
the ship off the rocks, no compressed air was used in the after hold to
lift the ship. The anchors and cables were in both cases transferred
aft, for the same purpose, namely, to diminish the upward pressure
forward. In the case of the wooden deck leaking, 200 of the same casks
were placed between it and the lower deck in the foremost hold to retain
some of the buoyancy of the forepart, which would otherwise be lost. No
decks were built in the compartment before the collision bulkhead, as
very little buoyancy was lost by that space being full of water, and all
that was there was confined to that compartment by the bulkhead and the
iron lower deck.

While all these foregoing arrangements were being made for the exclusion
of water from the inside of the ship, the engineers and firemen were
employed clearing the engine room of some fifty tons of coal which had
been washed from the open bunkers into the machinery by the sea, when
the engine room was full and the ship on the reef. The greatest
difficulty was experienced in digging out and excavating the engines
from the coal and dirt, and still greater was the labor of cleaning all
the mechanism and putting everything once more in an efficient steaming
condition. But all was finished soon after the decks had been completed,
and on October 12 she was ready for sea. On the following day she was
floated off and started on her perilous voyage to Halifax, using her own
engines, and making about five and a half knots an hour. Her steam pumps
were by this time all ready for service to assist the big ones on deck
in an emergency. She anchored once on her way round, at Shelburne, on
the coast of Nova Scotia, arriving at Halifax at 1 p.m. on October 17.
The trip round was a very anxious time for all hands, more especially
when they were overtaken by a fresh gale in the Atlantic, for the
forward deck was very liable to be burst up with the increased pressure
on it caused by the pitching of the ship; also the rudder was entirely
unable to bear any strain on it, because the lower part of the rudder
post was unconnected with the stern post, part of the stern framing
which connects the two having been broken off. Any heavy sea was
therefore likely to carry away the rudder altogether, or the same
accident might happen if the helm was put down too hard, rendering the
ship unmanageable.

She was placed in dry dock as soon as she arrived at Halifax, and it was
not until then that the full extent of the damages, caused by the
pounding on the rocks, could be fully realized. The first 20 feet of the
keel had been torn completely out, and about 30 feet from the stern
there was an immense hole, with the thick plates torn and bent like
paper, the framing and stanchions being twisted into all sorts of shapes
almost beyond recognition. Under the foremast the bottom of the ship was
bent up in the form of an arch, having been raised 4 feet above its
natural position, with an immense hole punctured on the starboard side,
besides several smaller ones. Also the aftermost 20 feet of keel was
torn and jagged, with several small holes in the skin, and the lower
portion of the stern framing was broken off, leaving the rudder post to
hang down unsupported at its extremity. It would strike one on looking
at these gaping wounds that it would be nearly impossible to place the
Ulunda in an efficient condition again, but the work of renewing the
damaged plates is being carried out at a great rate, and in three
months' time it is hoped that all the repairs necessary will be
completed and the ship once more doing her duty. She has already cost
her owners some $10,000, and $40,000 are estimated to cover all future

The foremast was snapped off in a somewhat novel manner. She was pivoted
on the rocks by her bows, and at high tide, the day after she struck, a
breeze sprang up and turned her round; the tide sinking again, the whole
weight of the ship came on the bottom of the ship where she was then
touching, namely, just on the spot where the foremost was stepped, and
right astern, leaving the center portion of the ship unsupported. This
caused the foremast to rise, and it being held down by wire rigging, it
snapped in several places, at the same time tearing up the shrouds from
the deck. This accounts also for the arch-like bulge in the bottom at
that spot and for the damages astern; also for the fact that Captain
Kelly discovered the ship with her head to sea.

Another incident happened when the ship was just rising off the rocks,
which nearly resulted in a catastrophe. When the ship was just beginning
to lift, the leak in No. 3 compartment was found to be gaining on its
pump. A diver was at once sent down to ascertain the cause, and he found
that a small hole, about 6 inches square, had been punctured in the
skin, which until then had been kept tight by the rock that had caused
it. It was necessary to close this leak at once. An iron bolt, which was
screwed for a nut at one end, was obtained and passed through a strong
piece of wood about 2 feet square. The inside of this board was
cushioned with canvas and oakum, and it was taken down outside the ship
by the diver and placed over the hole, with the feathered end of the
bolt sticking through the hole; the diver was then sent down inside the
hold, and with a nut set up the whole cushion until the flow of water
was stopped. The leak was thus stopped which had threatened the
arrangements for floating the ship with failure.

It has been seen that the method of raising the Ulunda was very simple.
She was floated off by the rising tide. If there had been only a small
instead of an 18 foot rise, some other mode would have to have been
adopted. No attempt was made to stop any of the leaks, except the one
just stated, but a deck above the lacerations was made water-tight, and
this, together with the sides of the ship hanging down, formed a kind of
diving bell, the pressure of air in which, caused by the water outside,
acting on this deck, being the principal means of buoying up the ship,
assisted by the buoyancy of the two water-tight compartments. The deck
afterward built was only necessary for the safety of the ship, she being
able to float without it; but it would have been suicidal to trust the
ship on the Atlantic in the state she was in when raised, since with any
swell on, the compressed air would escape and its place be taken by
water, the buoyancy necessary for keeping her afloat being thus lost.

It only remains to be said that the risks run in steaming around to
Halifax by herself were, as it was, very great, and had the wind and sea
been less favorable, the undertaking would probably have proved a
disastrous failure.

* * * * *


Up to recent years there was no reason for putting the question that
forms the title of this article, for it was admitted by all that the
conversion of thermic energy, or heat produced by the combustion of
coal, into mechanical energy or work could no longer be effected
economically except by having recourse to steam. In ordinary language,
and even to manufacturers, steam engine was the equivalent of thermic
motor, and it would not have occurred to any one to use anything else
but steam to effect the transformation.

The progress that has been made during the last twenty years in the
thermatic study and construction of gas motors (without speaking of hot
air motors) has shown that the use of steam is not absolutely
indispensable for the production of work, and it has demonstrated that,
as regards dynamic product, the gas motor preserves the advantage,
although the relatively high price of the illuminating gas employed in
the production of the motive power generally renders the use of this
combustible more costly than steam, especially for high powers.

The economic truth of twenty years ago, when gas motors absorbed 1,500
liters per horse hour and exceeded with difficulty an effective power of
from 8 to 10 horses, has become less and less certain, when the
consumption has successively descended to 1,200, 1,000, 800 and even to
600 liters of gas per horse hour, the power of the motors rising
gradually to 25, 50 and 100 horses with a motor having a single cylinder
of a diameter of 57 centimeters.


A, cylinder; B, condenser; C, boiler; R, feed water heater; D, chimney.]

But these results did not suffice, and it was desired to do better still
by dispensing with the use of high priced illuminating gas. An endeavor
was made to obviate the difficulty by manufacturing a special gas for
the motive power, as steam is produced for the same object, by
distilling coal, carbureting air, producing water gas by the Dowson
process, and by other equivalent processes.

The strides made in this direction were finally crowned with success,
and the results obtained in the recent experiments due to Mr. Aime Witz,
an undoubted authority in the matter, permit of affirming that now and
hereafter, in many circumstances, a gas generator supplying a gas motor
will be able to advantageously dethrone a steam boiler supplying a steam
engine of the same power.

These conclusions, which tend to nothing less than to limit the reign of
the steam engine, are confirmed on the one hand by an experiment carried
on for the last two years in the Barataud flour mill of Marseilles,
where a 50 h.p. "Simplex" motor has been running day and night for
several months without stopping, and consuming but about 500 grammes of
English anthracite per effective horse hour, and, on another hand, by
some personal experiments of Mr. Witz's, to which we shall shortly
advert, and whence there results a sensibly equivalent production for a
motor of 100 indicated h.p., corresponding to a power of 75 effective

Before establishing, with Mr. Witz, a comparison of the two systems in
pressure, steam or gas, let us state in a few words in what the latter
consists, the steam engine and the boiler that supplies it being so well
known that no description is necessary.

The Dowson gas generator does not differ essentially from the numerous
generators devised during recent years for the manufacture of gaseous
combustibles, the use of which is so often convenient. The motor that it
supplies is the most powerful single cylinder one that has hitherto been
constructed. It is of 100 indicated h.p., and its normal angular
velocity is 100 revolutions per minute. On trial it has yielded 112
indicated h.p., and 76.8 effective h.p., corresponding to an organic
rendering of 69 per cent. This motor, elaborated by Messrs.
Delamare-Bouteville & Malandin, of Rouen, operates by compression and in
four periods, according to the Beau de Rochas cycle. We give the aspect
of it in Fig. 3. In the first period the mixture of air and gas is
sucked in, in the second it is compressed, in the third it is ignited,
and in the fourth the products of combustion are expelled.


A, cylinder; B, gas conduit; C, rubber pockets; D. gasometer; E,
purifier; F, scrubber; G, hydraulic main; H, cooling pipe; I steam
injector; K, steam boiler and superheater; L, gas generator; M, charger;
N, discharge of the motor.]

Ignition is effected electrically by a series of sparks playing between
two platinum points in the slide valve, and this permits of regulating
the instant of ignition through the edges of the orifices. The angular
velocity is regulated by a Watt's governor, which secures an isochronism
of the motion independently of the charge.

The setting in motion of so powerful an engine is effected very easily
by means of an arrangement that permits of introducing into the
cylinder, while the piston is in the center of the stroke, a mixture of
air and gas whose pressure is sufficient at the arrival to expel the
inert products. After this the ignition takes place, and the explosion
is sufficient to set the motor in motion.

The trials made by Mr. Witz with the motor represented in Fig. 3 gave
the following results, deduced from an experiment of 68 hours. The
figures relate to one effective horse power, measured with the brake
upon the shaft of the motor.

Consumption of anthracite. 516 grammes.
" " coke. 96 "
Consumption of water for the injection
of steam. 0.487 liters.
Consumption of water for cooling
the cylinder. 50.0 "
Consumption of oil for lubricating
the cylinder. 3.74 grammes.
Consumption of grease. 0.45 "
Consumption of gas reduced to
0 deg. C. and to 760 mm. 2,370 liters.

This last figure will appear very high, but the fact must not be lost
sight of that it is a question of poor gas, the net cost of which varies
between one and two centimes per cubic meter, and the calorific power of
which is but 1,487 heat units per cubic meter of constant volume, and
supposing the steam condensed. This combustion of 612 grammes of
combustible per effective horse hour is remarkable, and fully shows what
may be expected of the gas motor supplied by a gas generator in putting
to profit certain improvements that will hereafter be possible, such,
for example, as the lightening of the movable parts of the motor, the
bettering of its organic rendering (now quite feeble), the use of better
oils, the reduction of the consumption of water, the superheating of the
steam injected into the gas generator, etc.

A well constructed steam engine, carefully kept in repair and as much
improved as it is possible to make it, would certainly consume twice as
much coal to produce the same quantity of effective work, say at least
1,200 grammes per horse hour. But, as has been objected with reason, it
does not suffice to compare the figures as to the consumption of fuel in
order to institute a serious comparison between the steam engine and the
motor using poor gas.

The gas generator requires the use of English anthracite, while a steam
boiler is heated with any kind of coal. The prices of unity of weight
are therefore very different. Moreover, the gas motor necessitates an
immense amount of water for the washing of the gas and the cooling of
the cylinder, through circulation in the jacket. It is well to keep this
fact in view. On another hand, the lubrification of the cylinders
requires a profusion of oil whose flashing point must be at a very high
temperature, else it would burn at every explosion and fill the cylinder
with coom. Such oil is very costly.

Does not the expenditure of oil in large motors largely offset the
saving in coal? And then, gas motors are sold at high prices, as are gas
generators, and this installation necessarily requires the addition of a
large gasometer, scrubbers, etc. The wear of these apparatus is rapid,
and if we take into account the interest and amortization of the capital
engaged, we shall find that the use of steam is still more economical.
The obstruction caused by bulky apparatus is another inconvenience, upon
which it is unnecessary to dwell. In a word, the question is a very
complex one. We look at but one side of it in occupying ourselves only
with the coal consumed, and we shall certainly expose those who allowed
themselves to be influenced by the seductive figures of consumption to
bitter disappointment.

To answer such objections Mr. Aime Witz has established a complete
parallel between the two systems, in which he looks at the question from
a theoretical and practical and scientific and financial point of view.
Considered as a transformation apparatus, a steam motor burning good
Cardiff coal in a Galloway boiler with feed water heaters will consume
(with a good condensing engine utilizing an expansion of a sixth) from
1,100 to 1,250 grammes of coal per effective horse hour, which
corresponds to a rough coefficient of utilization of 9.7 per cent. A gas
generator supplying a gas motor burning Swansea anthracite and Noeux
coke, medium quality, will consume 516 grammes of anthracite and 90 of
coke to produce 2,370 liters of gas giving 1,487 heat units per cubic
meter. Of the 3,524 heat units furnished to the motor by the 2,370
liters of gas, the motor will convert 18 per cent. into disposable
mechanical work.

With the boiler, the gross rendering of the whole is 7 per cent. With
the gas generator it reaches 12.7 per cent. From a theoretical point of
view the advantage therefore rests with the gas generator and gas motor.
In order to compare the net cost of the units of work, from an
industrial point of view, it is necessary to form estimates of
installation, costs of keeping in repair, interest and amortization.

Figs. 1 and 2 represent, on the same scale, the installations necessary
in each of these systems. The legends indicate the names of the
different apparatus in each installation. The following table shows
that, as regards the surface occupied, the advantage is again with the
gas generator and gas motor:

Steam Engine. Gas Motor.
Surface covered. 85 sq. m. 72 sq. m.
Surface exposed. 33 " 43 "
--- ---
Total surface. 118 " 115 "

The estimates of installation formed by Mr. Witz set forth the expense
relative to the capital engaged exactly at the same figure of 32,000
francs for a motive power of 75 effective horses. The expenses of
keeping in repair, interest, etc., summed up, show that the cost per day
of 10 hours is 47.9 francs for the steam engine and 39.6 for the gas
motor, say a saving of 8.3 francs per day, or about 2,500 francs for a
year of 300 days' work.

The gas motor, therefore, effects a great saving, while at the same time
occupying less space, consuming less water and operating just as well.

With Mr. Witz we cheerfully admit all the advantages that he so clearly
establishes with his perfect competency in such matters, but there still
remain two points upon which we wish to be enlightened. Are not the
starting up, the operation and the keeping in repair of a gas generator
actually more complicated and more delicate than the same elements of a
steam engine? Does not the poor gas manufactured in a gas generator
present, from a hygienic point of view, danger sufficiently great to
proscribe the use of such apparatus in many circumstances?

Such are the points upon which we should like to be enlightened before
unreservedly sharing Mr. Witz's enthusiasm, which, however, is
justified, economically speaking, by the magnificent results of the
experiments made by the learned engineer.--_La Nature_.


* * * * *


We publish illustrations of a Thwaites suspension pneumatic power 1/2 cwt.
hammer of a new design, for planishing pipes and plates, for which we
are indebted to _Engineering_. As indicated in the perspective view
(Fig. 1) the mechanism is supported at the center of a cross girder
resting on two cast iron square pillars, box section, each bolted down
to the foundations by four 11/4 in. diameter bolts. The measurements of
these columns and girders are given in Figs. 2 and 3, the former an
elevation of the hammer and the latter a plan, partly in section, of the
cross girder, while Fig. 4 is a cross section showing the arrangements
for operating the hammer. In the center is a cast iron guide for working
the ram, the guide being extended on two sides to receive the disk crank
journals, 2 in. in diameter by 31/2 in. long. The disk cranks are
connected to a hollow steel ram by a connecting rod. The ram is divided
inside into two compartments, each having a phosphor bronze air piston.
These are connected together by a steel piston rod, the top air piston
forming a connection for the small end of the connecting rod. The
outside diameter of the ram is 33/4 in., and the diameter of the air
pistons 23/4 in. and 2-7/8 in. respectively. Cottered into the bottom of
the ram is a steel pallet holder with a dovetail, so that the pallet can
be renewed or exchanged for one of another shape when required. Keyed on
to the crankshaft is a flanged pulley 10 in. in diameter by 31/4 in.
between flanges. There is also an overhead countershaft with strap
shifting arrangement. At the side of one of the columns a hand lever and
quadrant are provided, as shown in the perspective view and in Fig. 2,
for working an arrangement for tightening the belt when the machine is
working. To this arrangement is connected a powerful brake which stops
the machine in a few revolutions. It will be seen that the brake is
applied as the belt is slackened for stopping the machine. For
planishing pipes or tubes a long wrought iron mandrel is provided
mounted on two cast iron carriages, each having four flanged wheels for
running on rails. The hammer is arranged so that tubes 4 feet in
diameter can be worked for planishing plates. A pallet is fastened on
the top of one of the mandrel carriages, Figs. 5 to 8 showing the
details of the carriages. The general dimensions are: Distance between
pillars, 6 feet; height under girder, 5 feet; height from ground to top
of mandrel, 4 feet 13/4 in.; and length of stroke, 5 in. This machine is
capable of delivering 500 blows per minute. The constructors are Messrs.
Thwaites Brothers, Limited, Bradford, Yorkshire.

[Illustration: FIG. 1.]


* * * * *




By WM. L. SAUNDERS, C.E., of New York.

I cannot but realize as I stand before you that I would be very much
more at home were I in your midst. I feel but little older and so very
much less wise than when I sat in the class room an undergraduate of the
University of Pennsylvania, that I trust I may expect you to give me
this afternoon, not only your attention, but your sympathy.

The present situation is not without suggestions of my own experience. I
recall a lecture in the ordinary course, given by our professor of
mining, whose struggles with the English language were quite as
conspicuous as were our efforts to tell what he was driving at. He was
describing an ordinary windlass hoist used at the shaft of a mine. He
said "There is a windlass at de top of de shaft around which is coiled a
rope, on de two ends of which is fastened two er--er--_pans_, one of
which is a _bucket_ and de oder a _platform_." I mention this because I
shall ask you to attribute my shortcomings in this lecture, not so much
to my lack of familiarity with my native tongue, as to--well, because I
was not educated at Cornell University.

We all know what free air is. You who are privileged to live upon these
beautiful hills overlooking Ithaca and the lake, doubtless know more
about free air than we do who are choked in the dusty confines of New
York City. Compressed air is simply air under pressure. That pressure
may be an active one, as in the case of the piston of an air compressor;
or passive, as with the walls of a receiver or transmission pipe. It is
usual to define compressed air as air increased in density by pressure,
but we know that we may produce compressed air by heat alone. A simple
illustration of this is the pressure which will blow a cork from an
empty bottle when that bottle has been placed near the fire. Here we
have pressure, or compressed air, in the bottle produced by heat alone.

Having defined compressed air, we must next define heat; for in dealing
with compressed air, we are brought face to face with the complex laws
of Thermodynamics. We cannot produce compressed air without also
producing heat, and we cannot use compressed air as a power without
producing cold. Based on the material theory of heat, we would say that
when we take a certain volume of free air and compress it into a smaller
space, we get an increase in temperature because we have the heat of one
volume occupying less space, but no one at this date accepts the
material theory of heat. Your distinguished director, Professor
Thurston, in discussing "Steam and its Rivals," in the _Forum_, said:
"The science of Thermodynamics teaches that heat and mechanical energy
are only different phases of the same thing, the one being the motion of
molecules, and the other that of masses." This is the accepted theory of
heat. In other words, we do not believe that there is any such _thing_
as heat, but that what we call heat is only the sensible effect of
motion. In the cylinder of an air compressor the energy of the piston is
converted into molecular motion in the air and the result, or the
equivalent, is heat. A higher temperature means an increased speed of
vibration, and a lower temperature means that this speed of vibration is
reduced. If I hold an open cylinder in my left hand and a piston in my
right, and place the piston within the cylinder, I here have a confined
volume of air at the temperature and the pressure of this room. These
particles of air are in motion and produce heat and pressure in
proportion to that motion. Now if I press the piston to a point in the
center of the cylinder, that is, to one-half the stroke, I here decrease
the distance between the cylinder head and the piston just one-half,
hence each molecule of air strikes twice as many blows upon the piston
and head in traveling the same distance and the pressure is doubled. We
have also produced about 116 degrees of heat, because we have expended a
certain amount of work upon the air; the air has done no work in return,
but we have increased the energy of molecular vibration in the air and
the result is heat.

But what of this heat? What harm does it do? If I instantly release the
piston which I hold at one-half stroke it will return to its original
position, less only a little friction. I have, therefore, recovered all,
or nearly all, the power spent in compressing the air. I have simply
pressed a spring, and have let it recover. We see what a perfect spring
compressed air is. We see the possibility of expending one horse power
of energy upon air and getting almost exactly one horse power in return.
Such would be the case provided we used the compressed air power
_immediately and at the point where the compression takes place_. This
is never done, but the heat which has been boxed up[1] in the air is
lost by radiation, and we have lost power. Let us see to what extent
this takes place.

[Footnote 1: I use material terms because they add to simplicity of
expression and notwithstanding the fact that heat is vibration.]

Thirteen cubic feet of free air at normal temperature and barometric
pressure weigh about one pound. We have seen that 116 degrees of heat
have been liberated at half stroke. The gauge pressure at this point
reaches 24 pounds. According to Mariotte's law, "The temperature
remaining constant, the volume varies inversely as the pressure," we
should have 15 pounds gauge pressure. The difference, 9 pounds,
represents the effect of the heat of compression in increasing the
relative volume of the air.


The specific heat of air under constant pressure being 0.238, we have
0.238 x 116 = 27.6 heat units produced by compressing one pound or
thirteen cubic feet of free air into one-half its volume. 27.6 x 772
(Joule's equivalent) = 21,307 foot pounds. We know that 33,000 foot
pounds is one horse power, and we see how easily about two-thirds of a
horse power in heat units may be produced and lost in compressing one
pound of air. I would mention here that exactly this same loss is
suffered when compressed air does work in an engine and is expanded down
to its original pressure. In other words, _the heat of compression and
the cold of expansion are in degree equal_.

Experiments made by M. Regnault and others on the influence of heat on
pressures and volumes of gases have enabled us to fix the absolute zero
of temperature as -461 degrees Fahrenheit. This point, 461 degrees below
zero, is the theoretical point at which a volume of air is reduced to
nothing. The volume of air at different temperatures is in proportion to
the absolute temperature, and on this basis Box gives us the following


Temperature Volume in Weight of a
in degrees. cubic feet. cubic foot in lb.
32 1.000 0.0807
42 1.020 0.0791
52 1.041 0.0776
62 1.061 0.0761
72 1.082 0.0747
82 1.102 0.0733
92 1.122 0.0720
102 1.143 0.0707
112 1.163 0.0694
122 1.184 0.0682
132 1.204 0.0671
142 1.224 0.0660
152 1.245 0.0649
162 1.265 0.0638
172 1.285 0.0628
182 1.306 0.0618
192 1.326 0.0609
202 1.347 0.0600
212 1.367 0.0591
230 1.404 0.0575
250 1.444 0.0559
275 1.495 0.0540
300 1.546 0.0522
325 1.597 0.0506
350 1.648 0.0490
375 1.689 0.0477
400 1.750 0.0461
450 1.852 0.0436
500 1.954 0.0413
550 2.056 0.0384
600 2.15[1] 0.0376
650 2.260 0.0357
700 2.362 0.0338
750 2.464 0.0328
800 2.566 0.0315
850 2.668 0.0303
900 2.770 0.0292
950 2.872 0.0281
1,000 2.974 0.0268
1,100 3.177 0.0254
1,200 3.381 0.0239
1,300 3.585 0.0225
1,400 3.789 0.0213
1,500 3.993 0.0202
1,600 4.197 0.0192
1,700 4.401 0.0183
1,800 4.605 0.0175
1,900 4.809 0.0168
2,000 5.012 0.0161
2,100 5.216 0.0155
2,200 5.420 0.0149
2,300 5.624 0.0142
2,400 5.828 0.0138
2,500 6.032 0.0133
2,600 6.236 0.0130
2,700 6.440 0.0125
2,800 6.644 0.0121
2,900 6.847 0.0118
3,000 7.051 0.0114
3,100 7.255 0.0111
3,200 7.459 0.0108

[Transcribers note 1: last digit illegible]

The effect of this heat of compression in increasing the volume, and the
heat produced at different stages of compression, are shown by the
following table:


| Pressure. | | |
Atmo- +-----------+-----------+ Volume |Temperature | Total
spheres.|Pounds per |Pounds per | in Cubic | of the Air | Increase of
|Square Inch|Square Inch| Feet. | throughout | Temperature.
| above a |above the | |the Process.| Degrees.
| Vacuum. |Atmosphere | | Degrees. |
| |(Gauge | | |
| |Pressure). | | |
1.00 | 14.70 | 0.00 | 1.0000 | 60.0 | 00.0
1.10 | 16.17 | 1.47 | 0.9346 | 74.6 | 14.6
1.25 | 18.37 | 3.67 | 0.8536 | 94.8 | 34.8
1.50 | 22.05 | 7.35 | 0.7501 | 124.9 | 64.9
1.75 | 25.81 | 11.11 | 0.6724 | 151.6 | 91.6
2.00 | 29.40 | 14.70 | 0.6117 | 175.8 | 115.8
2.50 | 36.70 | 22.00 | 0.5221 | 218.3 | 158.3
3.00 | 44.10 | 29.40 | 0.4588 | 255.1 | 195.1
3.50 | 51.40 | 36.70 | 0.4113 | 287.8 | 227.8
4.00 | 58.80 | 44.10 | 0.3741 | 317.4 | 257.4
5.00 | 73.50 | 58.80 | 0.3194 | 369.4 | 309.4
6.00 | 88.20 | 73.50 | 0.2806 | 414.5 | 354.5
7.00 | 102.90 | 88.20 | 0.2516 | 454.5 | 394.5
8.00 | 117.60 | 102.90 | 0.2288 | 490.6 | 430.6
9.00 | 132.30 | 117.60 | 0.2105 | 523.7 | 463.4
10.00 | 147.00 | 132.30 | 0.1953 | 554.0 | 494.0
15.00 | 220.50 | 205.80 | 0.1465 | 681.0 | 621.0
20.00 | 294.00 | 279.30 | 0.1195 | 781.0 | 721.0
25.00 | 367.50 | 352.80 | 0.1020 | 864.0 | 804.0

A cubic foot of free air at a pressure of one atmosphere (equal to 14.7
pounds above a vacuum) at a temperature of 60 degrees, when compressed
to twenty-five atmospheres, will register 367.5 pounds above a vacuum
(352.8 pounds gauge pressure), will occupy a volume of 0.1020 cubic
foot, will have a temperature of 864 degrees, and the total increase of
temperature is 804 degrees.

The thermal results of air compression and expansion are shown by the
accompanying diagram.

The horizontal and vertical lines are the measures of volumes, pressures
and temperatures. The figures at the top indicate pressures in
atmospheres above a vacuum, the corresponding figures at the bottom
denote pressures by the gauge. At the right are volumes from one to
one-tenth. At the left are degrees of temperatures from zero to 1,000
Fahrenheit. The two curves which begin at the upper left hand corner and
extend to the lower right are the lines of compression or expansion.

The upper one being the _Adiabatic_ curve, or that which represents the
pressure at any point on the stroke with the heat developed by
compression remaining in the air; the lower is the _Isothermal_, or the
pressure curve uninfluenced by heat. The three curves which begin at the
lower left hand corner and rise to the right are heat curves and
represent the increase of temperature corresponding with different
pressures and volumes, assuming in one case that the temperature of the
air before admission to the compressor is zero, in another sixty
degrees, and in another one hundred degrees.

Beginning with the adiabatic curve, we find that for one volume of air
when compressed without cooling the curve intersects the first vertical
line at a point between 0.6 and 0.7 volume, the gauge pressure being
14.7 pounds. If we assume that this air was admitted to the compressor
at a temperature of zero, it will reach about 100 degrees when the gauge
pressure is 14.7 pounds. We find this by following down the first line
intersected by the adiabatic curve to the point where the zero heat
curve intersects this same line, the reading being given in figures to
the left immediately opposite. If the air had been admitted to the
compressor at 60 degrees, it would register about 176 degrees at 14.7
pounds gauge pressure. If the air were 100 degrees before compression,
it would go up to about 230 degrees at this pressure. Following this
adiabatic curve until it intersects line No. 5, representing a pressure
of five atmospheres above a vacuum (58.8 lb. gauge pressure), we see
that the total increase of temperature on the zero heat curve is about
270 degrees, for the 60 degree curve it is about 370 degrees, and for
the 100 degree curve it is about 435 degrees. The diagram shows that
when a volume of air is compressed adiabatically to 21 atmospheres (294
lb. gauge pressure), it will occupy a volume a little more than
one-tenth; the total increase of temperature with an initial temperature
of zero is about 650 degrees; with 60 degrees initial temperature it is
800 degrees, and with 100 degrees initial it is 900 degrees. It will be
observed that the zero heat curve is flatter than the others, indicating
that when free air is admitted to a compressor cold, the relative
increase of temperature is less than when the air is hot. This points to
the importance of low initial temperature.

We have now seen that the economical production of compressed air
depends upon the following conditions:

(1) A low initial temperature.

(2) Thorough cooling during compression.

It has been demonstrated by experiments made in France that the power
required to compress moist air is less than that for dry air. A table
showing the power required to compress moist and dry air has been
prepared from the data of M. Mallard and shows that for five atmospheres
the work expended in compressing one pound of dry air is 58,500 foot
pounds, while that for moist air is 52,500 foot pounds. In expansion
also moisture in the air adds to the economy, but in both cases the
saving of power is not great enough to compensate for the many
disadvantages due to the presence of water. Mr. Norman Selfe, of the
Engineering Association of N.S.W., has compiled a table which shows some
important theoretical conditions involved in producing compressed air.

So much for the theory of compression. We now come to the practical
production of compressed air.

The first record that we have of the use of an air compressor is at
Ramsgate Harbor, Kent, in the year 1788. Smeaton invented this "pump"
for use in a diving apparatus. In 1851, William Cubitt, at Rochester
Bridge, and a little later an engineer, Brunel, at Saltash, used
compressed air for bridge work. But the first notable application of
compressed air is due to Professor Colladon, of Geneva, whose plans were
adopted at the Mont Cenis tunnel. M. Sommeiller developed the Colladon
idea and constructed the compressed air plant illustrated in Fig. 2.

[Illustration: FIG. 2.]

The Sommeiller compressor was operated as a ram, utilizing a natural
head of water to force air at 80 pounds pressure into a receiver. The
column of water contained in the long pipe on the side of the hill was
started and stopped automatically, by valves controlled by engines. The
weight and momentum of the water forced a volume of air with such shock
against a discharge valve that it was opened and the air was discharged
into the tank; the valve was then closed, the water checked; a portion
of it was allowed to discharge and the space was filled with air, which
was in turn forced into the tank. The efficiency of this compressor was
about 50 per cent.

At the St. Gothard tunnel, begun in 1872, Prof. Colladon first
introduced the injection of water in the form of spray into the
compressor cylinder to absorb the heat of compression.

[Illustration: FIG. 3.]

Fig. 3 illustrates the air cylinder of the Dubois-Francois type of
compressor, which was the best in use about the year 1876. This
compressor was exhibited at the Centennial Exposition and was adopted by
Mr. Sutro in the construction of the Sutro tunnel. A characteristic
feature seems to be to get as much water into the cylinder as possible.
The water which flooded the bottom of the cylinder arose from the
voluminous injection; this water was pushed into the end of the cylinder
and some of it escaped with the air through the discharge valve.

An improved pattern of this compressor is shown in Fig. 4.

[Illustration: FIG. 4.]

These illustrations are interesting from an historical point of view, as
indicating the line of thought which early designers of air-compressing
machinery followed. As the necessity for compressed air power grew,
inventors turned their attention to the construction of air-compressing
engines that would combine _efficiency_ with _light weight_ and _economy
of space and cost_. The trade demanded compressors at inaccessible
localities, and in many cases it was preferred to sacrifice isothermal
results to simplicity of construction and low cost.

It is evident that an air compressor which has the steam cylinder and
the air cylinder on a single straight rod will apply the power in the
most direct manner, and will involve the simplest mechanics in the
construction of its parts. It is evident, however, that this straight
line, or direct construction, results in an engine which has the
greatest power at a time when there is no work to perform. At the
beginning of the stroke steam at the boiler pressure is admitted behind
the piston, and, as the air piston at that time is also at the initial
point in the stroke, it has only free air against it. The two pistons
move simultaneously, and the resistance in the air cylinder rapidly
increases as the air is compressed. To get economical results it is, of
course, necessary to cut off in the steam cylinder, so that at the end
of the stroke, when the steam pressure is low, as indicated by the
dotted line (Fig. 5), the air pressure is high, as similarly indicated.
The early direct-acting compressor used steam at full pressure
throughout the stroke. The Westinghouse pump, applied to locomotives, is
built on this principle, and those who have observed it work have
perhaps noticed that its speed of stroke is not uniform, but that it
moves rapidly at the beginning, gradually reducing its speed, and seems
to labor, until the direction of stroke is reversed. This construction
is admitted to be wasteful, but in some cases, notably that of the
Westinghouse pump, economy in steam consumption is sacrificed to
lightness and economy of space.

[Illustration: FIG. 5.]

Many efforts were made to equalize the power and resistance by
constructing the air compressor on the crank shaft principle, putting
the cranks at various angles, and by angular positions of steam and air
cylinders. Several types are shown in Fig. 6.

[Illustration: FIG. 6.]

Angular positions of the cylinder involve expensive construction and
unsteadiness. Experience has conclusively proved that it does not pay to
build air compressors with vertical cylinders, and moreover we have
found out that there is nothing in the apparent difficulty in equalizing
the strains in a direct-acting engine. It is simply necessary to add
enough weight to the moving parts, that is, to the piston, piston rod,
fly wheel, etc., to cut off early in the stroke and secure rotative
speed with the most economical results and with the cheapest
construction. It is obvious that the theoretically perfect air
compressor is a direct-acting one with a conical air cylinder, the base
of the cone being nearest the steam cylinder. This, from a practical
point of view, is impossible. Mr. Hill, in referring to the fallacious
tendencies of pneumatic engineers to equalize power and resistance in
air compressors, says: "The ingenuity of mechanics has been taxed and a
great variety of devices have been employed. It is usual to build on the
pattern of presses which do their work in a few inches of the end of the
stroke and employ heavy fly wheels, extra strong connections, and
prodigious bed plates. Counterpoise weights are also attached to such
machines; the steam is allowed to follow full stroke, steam cylinders
are placed at awkward angles to the air-compressing cylinders and the
motion conveyed through yokes, toggles, levers; and many joints and
other devices are used, many of which are entire failures, while some
are used with questionable engineering skill and very poor results."

[Illustration: FIG. 7.]

Fig. 7 illustrates the theory of Duplex Air Compressors. The hydraulic
piston or plunger compressor is largely used in Germany and elsewhere on
the Continent of Europe, but the duplex may be said to be the standard
type of European compressor at the present time. It is also largely used
in this country. Fig. 7 shows the four cylinders of a duplex compressor
in two positions of the stroke. It will be observed that each steam
cylinder has an air cylinder connected directly to the tail rod of its
piston, so that it is a direct-acting machine, except in that the
strains are transmitted through a single fly wheel, which is attached to
a crank shaft connecting the engines. In other words, a duplex air
compressor would be identical with a duplex steam engine were it not for
the fact that air cylinders are connected to the steam piston rods. The
result is, as shown in Fig. 7, that, at that point of the stroke
indicated in the top section, the upper right hand steam cylinder,
having steam at full pressure behind its piston, is doing work through
the angle of the crank shaft upon the air in the lower left hand
cylinder. At this point of the stroke the opposite steam cylinder has a
reduced steam pressure and is doing little or no work, because the
opposite air cylinder is beginning its stroke. Referring now to the
lower section, it will be seen that the conditions are reversed. One
crank has turned the center, and that piston which in the upper section
was doing the greatest work is now doing little or nothing, while the
labor of the engine has been transferred to those cylinders which a
moment before had been doing no work.

There are some advantages in this duplex construction, and some
disadvantages. The crank shafts being set quartering, as is the usual
construction, the engine may be run at low speed without getting on the
center. Each half being complete in itself, it is possible to detach the
one when only half the capacity is required. The power and resistance
being equalized through opposite cylinders, large fly wheels are not
necessary. Strange to say, the American practice seems to be to attach
enormous fly wheels to duplex air compressors. It is difficult to
justify this apparently useless expense in view of the facts shown in
Fig. 7. A fly wheel does not furnish power, nor does it add to the
economy of an engine except in so far as it enables it to cut off early
in the stroke, and to equalize the power and resistance. In other words,
a fly wheel is not a _source_ of power, and in many cases it is only a
means by which we accomplish rotative speed. It takes power to move
matter, and assuming that other conditions are equal, every engine that
carries a fly wheel that is larger than is necessary consumes a certain
number of foot pounds in turning so much metal around through space.
Were it possible to cut off at the same point and rotate as positively
without a fly wheel, it would be done away with entirely. Some straight
line air compressors are so constructed that the momentum of the piston
and other moving parts is nearly sufficient to equalize the strains
without a fly wheel; but the fly wheel is there because it insures a
definite length of stroke, and because it enables us to operate
eccentrics and to regulate the speed of the engine uniformly.

Objections to the duplex construction are: The strains are indirect,
angular and intermittent. It is necessary therefore to largely increase
the strength of parts; to add a crank shaft of large diameter with
enormous bearings, and to build expensive and very secure foundations.
Should the foundations settle at any point, excessive strains will be
brought upon the bearings, resulting in friction and liability to
breakage. A steam engine meets with a resistance on its crank shaft that
is uniform throughout the stroke; while an air compressor is subject to
a heavy maximum strain at the end of the stroke, hence the importance of
direct straight line connection between power and resistance.

[Illustration: FIG. 8.]

The friction loss on a duplex compressor seldom gets lower than 15 per
cent., while straight line compressors show as low a loss as 5 per cent.
Fig. 8 illustrates the Rand Duplex Air Compressor, a machine largely
used in America, especially in the Lake Superior iron mines. Fig. 9
illustrates a Duplex Compound Condensing Corliss Air Compressor built by
the Ingersoll-Sergeant Drill Company. This is a compressor made of the
best type of Corliss engine, with air cylinders connected to the tail
rods of the steam cylinders. One of these machines, of about 400 horse
power capacity, is now at work furnishing compressed air power for the
Brightwood Street Railway in Washington, D.C. Fig. 10 illustrates the
Norwalk direct-acting straight line air compressor, with compound air
cylinder. The chief purpose of compounding is to reduce the maximum
strain. This construction also adds to isothermal economy. The large
cylinder to the left determines the capacity of the compressor, the air
being compressed first to a low pressure (ordinarily about 30 pounds per
square inch), afterward passing through an intercooler, by which its
temperature is reduced, and then it is compressed still higher, even to
5,000 pounds per square inch if desired. The terminal strain, which is
so severe in air compressors, is here considerably reduced, as in this
case it is only equal to the area of the initial air piston multiplied
by its low air pressure.

[Illustration: FIG. 9.]

Economical results are attained with this compressor at low cost of
construction. The fly wheels are small, and the bearings narrow, because
the maximum strain is less, and the momentum of the piston and other
moving parts is such that most of the high initial steam power is taken
up in starting these parts and is afterward given out at the end of the
stroke, when the steam pressure is low. The strains are direct, and
expensive foundations are not required. Fig. 11 illustrates the
Ingersoll-Sergeant Compound Straight Line Air Compressor. This differs
from the one just described chiefly in that it is single-acting, while
the other is double-acting.

[Illustration: FIG. 10.]

By single-acting is meant that the air cylinders compress their
respective volumes of air _once_ every revolution. The air is admitted
to the large cylinder through the piston, is compressed to about 30
pounds, and on the return stroke the pressure is raised to almost any
point required, and in proportion to the diameter of the smaller
cylinder. Though single-acting, the capacity of one of these compressors
is about equal to that of the double-acting machine of the same cost of
construction. The initial air cylinder is made large enough to
correspond with the capacity of the smaller double-acting cylinder. The
strains are equalized because the area of the large cylinder multiplied
by its low pressure is exactly equal to that of the small cylinder
multiplied by its high pressure. The maximum strains are reduced
considerably below those which exist in compressors that do not compound
the air.

[Illustration: FIG. 11.]

The advantage of the single-acting air cylinder over the double is that
it compresses a volume of free air only once every revolution, hence
there is a better chance to cool the air during compression. The
cylinders have time to impart to the water jackets the heat produced by
compression and are kept cooler. The large air head of the initial
cylinder is jacketed, also adding to isothermal economy.

[Illustration: FIG. 12.]

Fig. 12 illustrates the Ingersoll-Sergeant Piston Inlet Cold Air
Compressor. This a straight line direct-acting engine, with steam and
air pistons connected to a single rod through a crosshead which connects
with two fly wheels. The strains are direct and the power and resistance
are equalized by the inertia of the crosshead, piston, rods, and fly
wheels. The Meyer's adjustable cut-off is used on the steam cylinder.
The air cylinder is provided with a tail rod tube through which all the
air is admitted into the cylinder.


Fig. 13 illustrates an unloading device and regulator applied to the
Ingersoll-Sergeant compressor.

The purpose of this unloading device is to maintain a uniform air
pressure in the receiver and a uniform speed of engine, notwithstanding
the consumption of the air, and to do this without waste of power or
attention on the part of the engineer. A weighted valve of safety valve
pattern is attached to the air cylinder, and is connected with the air
receiver, and with a discharge valve on each end of the air cylinder,
also with a balanced throttle valve in the steam pipe. When the pressure
of the air gets above the desired point in the receiver, the valve is
lifted and the air is exhausted from behind the discharge valves, thus
letting the compressed air at full receiver pressure into the cylinder
at both ends, and balancing the engine. At the same instant the
compressed air is exhausted from the little piston connected with the
balanced steam valve and the steam is automatically throttled, so that
only enough steam is admitted to keep the engine turning around, or to
overcome the friction, no work being done.

[Illustration: FIG. 14.]

When the compressor is unloaded, it is evident that the function of the
air piston is merely to force the compressed air through the discharge
valves and passages from one end to the other until more compressed air
is required, this being indicated by a fall in the receiver pressure.
The weighted valve now closes and the small connecting pipes are
instantly filled with compressed air; the steam valve automatically
opens, and the compression goes on in the regular way. Another function
of this device is to prevent the compressor from stopping or getting on
the center. Direct-acting compressors are liable to center when doing
work at slow speed.


Fig. 15 illustrates the Ingersoll-Sergeant Air Cylinder and Piston.

Fig. 16 shows the piston inlet valve, situated at G in Fig. 15. Two of
these valves are placed in each piston of a double-acting air cylinder,
the piston being hollow and the free air being admitted through a
tail-rod pipe, letter E, Fig. 15. JJ are water jacket passages for
cooling the air during compression. Owing to the absence of inlet
valves, large water jackets are provided, not only around the cylinder
itself, but through the heads. As the heat of compression is greater
near the end of the stroke, the advantage of a cool head is manifest. H
H are the discharge valves through which the compressed air is forced.


The most interesting feature of this cylinder is the piston inlet valve.
It is evident that this valve being attached to the piston needs no
springs or other connections, but is opened and closed exactly at the
right time by its natural inertia. With only about 1/4 of an inch throw of
valve a large area is opened, through which the free air is drawn. The
valve is made of a single piece of composition metal and is practically
indestructible. Its construction is such that it fills the clearance
spaces to a greater extent than is usual in air compressors. A singular
feature is that indicator cards taken on these cylinders show a free air
line in some cases a little above the atmospheric line. Poppet valve
compressors almost invariably show a slight vacuum, due to several
causes, mainly the duty performed in compressing the springs of the
valves, but the vacuum is also influenced by insufficiency of valve
area, hot air cylinders, etc. This cylinder gives its full volume of
air, and apparently a little more at times, because the air is admitted
by a concentrated inlet in which free _air is always moving in one
direction_. After it has been started, the speed of the compressor is
such that the air attains a momentum due to its velocity and density;
this serves a useful purpose in piling up the free air in the cylinder
before the inlet valve closes on the return stroke.


Taken from a 16x18 Sergeant piston inlet air compressor, meyer's cut-off
at 3/10. Steam at 58 lb.; air pressure, 77 lb.; total engine friction, 5
per cent.]

Fig. 17 illustrates a combined steam and air indicator card taken from
one of these cylinders. It will be observed that with steam and air
cylinders equal in diameter and stroke, an air pressure of 77 pounds is
reached with a steam pressure of only 58 pounds. The reason for this is
plainly shown in the cards, their areas being nearly equal. What is made
up in the air card by high pressure is represented in the steam card by
greater volume. The indicated efficiency deduced from these cards is 95
per cent., that is, the area of the air card divided by the area of the
steam card, representing the resistance divided by the power, results in
95 per cent. While several cards have been taken on the cylinders
showing a loss by friction of only 5 per cent., yet on the average the
best practice shows a loss of 6 per cent. or an efficiency of 94 per
cent. This result indicates an almost perfect proportion between power
and resistance, and good workmanship in air-compressing machinery. It is
difficult to conceive an engine of this size being worked with a less
expenditure for friction than 5 or 6 per cent. Were it possible to
retain the heat which is in the air, and which is represented by the
space between the dotted isothermal curve and the actual curve, we might
attain high efficiency in using compressed air power, but it is evident
that the power represented by the area of this space will be lost by
radiation of heat before it is used in an engine situated several
hundred feet away.

These indicator cards show at a glance that heat is responsible for the
important air losses, and that so far as the design of the compressing
engine is concerned, we have attained a point very near perfection. All
the devices, past, present and future, on which inventors spend so much
time, and in the development of which capitalists are innocently
inveigled, _aim to save this six per cent. loss!_ We hear a good deal
about "Centrifugal Air Compressors," "Rotaries," "Plunger Pumps," etc.,
designs involving expensive complications without any heat advantage,
and which seem to be based upon the "iridescent dream" of a large loss
in the present method of compressing air. Here we have a simple engine,
compact and complete in itself, capable of high speed without injury,
constructed on the basis of our best steam engine practice, which
produces compressed air power at a loss of only six per cent.

Clearance is not taken into consideration in the foregoing figures, but
clearance is very much more of a _bete noir_ in theory than in practice.
The early designers, as shown in the "Dubois-Francois" illustrations,
Figs. 3 and 4, regarded clearance loss as a very serious matter. Even at
the present time some air compressor manufacturers admit water through
the inlet valves into the air cylinder, not so much for the purpose of
cooling as to fill up the clearance space. A long stroke involving
expensive construction is usually justified by the claim that a large
saving is effected by reduced clearance loss. Let us see what the effect
of this clearance is. Assuming that we have an air compressor which
shows an isothermal pressure line, there would be some loss of power due
to clearance space, because we would have a certain volume of air upon
which work was done and heat produced, that heat having been absorbed
and the air being retained in the cylinder and not serving any useful
purpose. But let us assume that we have a compressor which shows an
adiabatic pressure line. We now have the air in the clearance space
acting precisely as a spring, compressed at each stroke, retaining its
heat of compression, and giving it out against the air piston at the
point when the stroke is reversed. There is no loss of power in such a
case as this, but, on the contrary, the air spring is useful in
overcoming the inertia of the piston and moving parts. The best air
compressors give a result about midway between the isothermal and the
adiabatic, and the net loss of _power_ directly due to clearance is so
small as to be practically unworthy of consideration.

It must not be inferred from the preceding remarks that the designer of
an air compressor may neglect the question of clearance. On the
contrary, it is a very important consideration. If we assume a large
clearance space in the end of an air cylinder of a compressor which is
furnishing air at a high pressure, we may readily conceive that space to
be so large, and that pressure so high, that the entire volume of the
cylinder would be filled by the air from the clearance space alone, and
the compressor would take in no free air and would, of course, produce
no compressed air.

Loss in _capacity_ of air compressors by clearance is in direct
proportion to the pressure.

Owing to the loss of capacity by clearance space at high pressures, it
is important that compound air cylinders should be used for furnishing
air at high pressure. With compound air cylinders the air is compressed
to alternate stages of pressure in the different cylinders, and the
clearance loss is thus reduced because of the reduced density of the air
in the clearance spaces. In ordinary practice air compressors deliver
the air at less than 100 pounds pressure, so that with a properly
designed air cylinder the clearance space is so small that the capacity
of the compressor is not materially affected.

Two systems are in use by which the heat of compression is absorbed, and
the difference between one and the other is so distinct that air
compressors are usually divided into two classes (1) wet compressors,
(2) dry compressors.

A _wet_ compressor is that which introduces water directly into the air
cylinder during compression.

A _dry_ compressor is that which introduces no water into the air during

_Wet_ compressors may be subdivided into two classes.

(1) Those which inject water in the form of a spray into the cylinder
during compression.

(2) Those which use a water piston for forcing the air into confinement.

The injection of water into the cylinder is usually known as the
Colladon idea. Compressors built on this system have shown the highest
isothermal results, that is, by means of a finely divided spray of cold
water the heat of compression has been absorbed to a point where the
compressed air has been discharged at a temperature nearly equal to that
at which it was admitted to the cylinder. The advantages of water
injection during compression are as follows:

(1) Low temperature of air during compression.

(2) Increased volume of air per stroke, due to filling of clearance
spaces with water and to a cold air cylinder.

(3) Low temperature of air immediately after compression, thus
condensing moisture in the air receiver.

(4) Low temperature of cylinder and valves, thus maintaining packing,

(5) Economical results, due to compression of moist air (see table 3).


| | |
|Compression at |Compression |
|a Constant |with |
|Temperature. |Increase of |
|Mariotte's Law. |Temperature. |
| | | | | | | | | | | | | | |
1|0.1 | | | | | | 20 | 68 |1.0 | | | 68 | | |
2|0.5 | 7199|1468|0.612| 7932|1618| 85.5|186 |1.222| 733|0.092|111 |3.0|23500|22500
3|0.333|11356|2316|0.459|13360|2725|130.4|267 |1.375|2004|0.150|135.5|4.0|37000|35000
4|0.25 |14260|2909|0.374|17737|3618|165.6|330 |1.495|3477|0.196|153.5|4.8|48500|45000
5|0.200|16580|3383|0.320|21209|4326|195.3|384 |1.595|4629|0.213|167 |5.4|58500|52500
6|0.167|18475|3768|0.281|24310|4959|220.5|429 |1.681|5835|0.240|179 |6.0|67000|60000
7|0.143|20038|4087|0.252|27048|5517|243.2|470 |1.758|7040|0.260|190 |6.4|75000|66000
8|0.125|21422|4370|0.229|29518|6021|263.6|506.5|1.828|8096|0.274| | | |
9|0.111| | |0.210| | |282 |539.6|1.891| | | | | |
10|0.100| | |0.195| | |299 |570.2|1.950| | | | | |
| | | | | | | | | | | | | | |
1| 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14| 15 | 16

Column Heading
1 Tension in Atmospheres.
2 Volume.
3 Work of Compression. Cubic Meters in Kilogram-meters.
4 Work of Compression. Cubic Feet in Foot Pounds.
5 Volume.
6 Work of Compression. (Dry.) Cubic Meters in Kilogram-meters.
7 Work of Compression. (Dry.) Cubic Feet in Foot Pounds. Deduced from 6.
8 Temperatures. (Dry.) Cent.
9 Temperatures. (Dry.) Fah.
10 Ratio of Greater to Less Temperature. Absolute.
11 Loss of Work in Compressing one Cubic Meter in Kilogram-meters.
By Increase of Temperature alone.
12 Percentage of Work of Compression Converted into Heat and Lost.
By Increase of Temperature alone.
13 Final Temperature if Water is used in Compression. Fah.
14 Percentage of Water to Air Required.
15 Foot Pounds to Compress One Pound Air. Dry.
16 Foot Pounds to Compress One Pound Air. With sufficient Moisture.

The first advantage is by far the most important one, and is really the
only excuse for water injection in air compressors. We have seen (table
3) that the percentage of work of compression which is converted into
heat and loss when no cooling system is used is as follows:

Compressing to 2 atmospheres loss 9.2 per cent.
" " 3 " " 15.0 " "
" " 4 " " 19.6 " "
" " 5 " " 21.3 " "
" " 6 " " 24.0 " "
" " 7 " " 26.0 " "
" " 8 " " 27.4 " "

We see that in compressing air to five atmospheres, which is the usual
practice, the heat loss is 21.3 per cent., so that if we keep down the
temperature of the air during compression to the isothermal line, we
save this loss. The best practice in America has brought this heat loss
down to 3.6 per cent. (old Ingersoll Injection Air Compressor), while in
Europe the heat loss has been reduced to 1.6 per cent. Steam-driven air
compressors are usually run at a piston speed of about 350 feet per
minute, or from 60-80 revolutions per minute of compressors of average
sizes, say 18" diameter of cylinder. Sixty revolutions per minute is
equal to 120 strokes, or two strokes per second. An air cylinder 18" in
diameter filled with free air once every half second, and at each stroke
compressing the air to 60 pounds, and thereby producing 309 degrees of
heat, is thus, by means of water injection, cooled to an extent hardly
possible with mere surface contact. The specific heat of water being
about four times that of air, it readily takes up the heat of

A properly designed spray system must not be confused with the numerous
devices applied to air cylinders, by means of which water is introduced.
In some cases the water is merely drawn in through the inlet valves. In
others it passes through the center of the piston and rod, coming in
contact with the interior walls of the air cylinder between the packing
rings. Introducing water into the air cylinder in _any other way, except
in the form of a spray, has but little effect in cooling the air during
compression._ On the contrary, it is a most fallacious system, because
it introduces all the disadvantages of water injection without its
isothermal influence. Water, by mere surface contact with air, takes up
but little heat, while the air, having a chance to increase its
temperature, absorbs water through the affinity of air for moisture, and
thus carries over a volume of saturated hot air into the receiver and
pipes, which on cooling, as it always does in transit to the mine,
deposits its moisture and gives trouble through water and freezing. It
is, therefore, of much importance to bear in mind that unless water can
be introduced _during compression_ to such an extent as to _keep down
the temperature of the air in the cylinder_, it had better not be
introduced at all.

If too little water is introduced into an air cylinder during
compression, the result is warm, moist air, and if too much water is
used, it results in a surplus of power required to move a body of water
which renders no useful service. The following table deduced from
Zahner's formula gives the quantity of water which should be injected
per cubic foot of air compressed in order to keep the temperature down
to 104 degrees Fah.

| | |
| |Weight of water |Weight of water
| |to be injected at |to be injected at
|Heat units devel-|68 deg. Fah. to keep |68 deg. Fah. to keep
Compression |oped in 1 lb. |the temperature at|the temperature at
by atmosphere |free air by |104 deg. Fah. in lbs. |104 deg. Fah. in lbs. of
above a volume.|compression. |of water and per |water for 1 cubic
| |lb. of free air. |foot of free air.
| | |
2 | 3.702 | 0.734 | 0.056
3 | 5.867 | 1.664 | 0.089
4 | 7.406 | 1.469 | 0.113
5 | 8.598 | 1.701 | 0.131
6 | 9.570 | 1.891 | 0.145
7 | 10.398 | 2.063 | 0.158
8 | 11.109 | 2.204 | 0.167
9 | 11.740 | 2.329 | 0.179
10 | 12.301 | 2.440 | 0.188
11 | 12.813 | 2.542 | 0.195
12 | 13.278 | 2.634 | 0.202
13 | 13.706 | 2.719 | 0.209
14 | 14.102 | 2.798 | 0.215
15 | 14.471 | 2.871 | 0.223

Objections to water injection are as follows:

(1) Impurities in the water, which, through both mechanical and chemical
action, destroy exposed metallic surfaces.

(2) Wear of cylinder, piston and other parts, due directly to the fact
that water is a bad lubricant, and as the density of water is greater
than that of oil, the latter floats on the water and has no chance to
lubricate the moving parts.

(3) Wet air arising from insufficient quantity of water and from
inefficient means of ejection.

(4) Mechanical complications connected with the water pump, and the
difficulties in the way of proportioning the volume of water and its
temperature to the volume, temperature and pressure of the air.

(5) Loss of power required to overcome the inertia of the water.

(6) Limitations to the speed of the compressor, because of the liability
to break the cylinder head joint by water confined in the clearance

(7) Absorption of air by water.

Before the introduction of condensing air receivers, wet air resulting
in freezing was considered the most serious obstacle to water injection;
but this difficulty no longer exists, as experience has conclusively
demonstrated that a large part of the moisture in compressed air may be
abstracted in the air receiver. Even in the so-called dry compressors a
great deal of moisture is carried over with the compressed air, because
the atmosphere is never free from moisture. This subject will be
referred to more fully when treating of the transmission of compressed

By far the most serious obstacle to water injection, and that which
condemns the wet compressor, is the influence of the injected water upon
the air cylinder and parts. Even when pure water is used, the cylinders
wear to such an extent as to produce leakage and to require reboring.
The limitation to the speed of a compressor is also an important
objection. The claim made by some that the injected water does not fill
the clearance spaces, but is aerated, does not hold good, except with an
inefficient injection system. The writer has increased the speed of an
air compressor (cylinders 12 in. and 12 in. by 18 in., injection air
cylinder) ten revolutions per minute by placing his fingers over the
orifice of the suction pipe of the water pump. The boiler pressure
remained the same, the cut-off was not changed and the air pressure was
uniform, hence this increase of speed arose from the fact that the water
was restricted and the clearance spaces were filled with compressed air,
which served as a cushion or spring. While the volume of compressed air
furnished by this compressor would be somewhat reduced by the
restriction of the water, yet the increase in speed which was obtained
without any increase of power fully compensated for the clearance loss.

Mr. John Darlington, of England, gives the following particulars of a
modern air compressor of European type:

"Engine, two vertical cylinders, steam jacketed, with Meyer's expansion
gear. Cylinders, 16.9 inches diameter, stroke 39.4 inches; compressor,
two cylinders, diameter of piston, 23.0 inches; stroke 39.4 inches;
revolutions per minute, 30 to 40; piston speed 39 to 52 inches per
second, capacity of cylinder per revolution, 20 cubic feet: diameter of
valves, viz., four inlet and four outlet, 51/2 inches; weight of each
inlet valve, 8 lb.; outlet, 10 lb.; pressure of air, 4 to 5 atmospheres.
The diagrams taken of the engine and compressor show that the work
expended in compressing one cubic meter of air to 4.21 effective
atmospheres was 38,128 lb. According to Boyle and Mariotte's law it
would be 37,534 lb., the difference being 594 lb., or a loss of 1.6 per
cent. Or if compressed without abstraction of heat, the work expended
would in that case have been 48,158. The volume of air compressed per
revolution was 0.5654 cubic meter. For obtaining this measure of
compressed air, the work expended was 21,557 pounds. The work done in
the steam cylinders, from indicator diagrams, is shown to have been
25,205 pounds, the useful effect being 851/2 per cent. of the power
expended. The temperature of air on entering the cylinder was 50 degrees
Fah., on leaving 62 degrees Fah., or an increase of 12 degrees Fah.
Without the water jacket and water injection for cooling the temperature
it would have been 302 degrees Fah. The water injected into the
cylinders per revolution was 0.81 gallon."

We have in the foregoing a remarkable isothermal result. The heat of
compression is so thoroughly absorbed that the thermal loss is only 1.6
per cent.; but the loss _by friction of the engine_ is 14.5 per cent.,
and the net economy of the whole system is no greater than that of the
best American dry compressor, which loses about one-half the theoretical
loss due to heat of compression, but which makes up the difference by a
low friction loss.

The wet compressor of the second class is the water piston compressor,
Fig. 18.


The illustration shows the general type of this compressor, though it
has been subject to much modification in different places. In America, a
plunger is used instead of a piston, and as it always moves in water the
result is more satisfactory. The piston, or plunger, moves horizontally
in the lower part of a U shaped cylinder. Water at all times surrounds
the piston, and fills alternately the upper chambers. The free air is
admitted through a valve on the side of each column and is discharged
through the top. The movement of the piston causes the water to rise on
one side and fall on the other. As the water falls the space is occupied
by free air, which is compressed when the motion of the piston is
reversed, and the water column raised. The discharge valve is so
proportioned that some of the water is carried out after the air has
been discharged. Hence there are no clearance losses.

This hydraulic compressor seems to have a certain charm about it, which
has resulted in its adoption in Germany, France and Belgium, and by one
of the largest mines in the United States. Its advantages are _purely
theoretical_, and without certain adjuncts which have been in some cases
applied to it, even the _theory_ is a very bad one.

The chief claim for this water piston compressor is that its piston is
also its cooling device, and that the heat of compression is absorbed by
the water. So much confidence seems to be placed in the isothermal
features of this machine that usually no water jacket or spray pump is
applied. Mr. Darlington, who is one of the stanch defenders of this
class of compressors, has found it necessary to introduce "spray jets of
water immediately under the outlet valves," the object of which is to
absorb a larger amount of heat than would otherwise be effected by the
simple contact of the air with the water-compressing column. Without
such spray connections, it is safe to say that this compressor has
scarcely any cooling advantages at all, so far as air cooling is
concerned. Water is not a good conductor of heat. In this case only one
side of a large body of air is exposed to a water surface, and as water
is a bad conductor, the result is that a thin film of water gets hot in
the early stage of the stroke and little or no cooling takes place
thereafter. The compressed air is doubtless cooled before it gets even
as far as the receiver, because so much water is tumbled over into the
pipes with it, but to produce economical results the cooling should take
place _during compression_.

Water and cast iron have about the same relative capacity for heat at
equal volumes. In this water piston compressor we have only one cooling
surface, which soon gets hot, while with a dry compressor, with water
jacketed cylinders and heads, there are several cold metallic surfaces
exposed on one side to the heat of compression, and on the other to a
moving body of cold water.

But the water piston fraternity promptly brings forward the question of
speed. They say that, admitting that the cooling surfaces are equal, we
have in one case _more time_ to absorb the heat than in the other. This
is true, and here we come to an important class division in air
compressing machinery--_high speed and short stroke_ as against _slow
speed and long stroke_. Hydraulic piston compressors are subject to the
laws that govern piston pumps, and are, therefore, limited to a piston
speed of about 100 feet per minute. It is quite out of the question to
run them at much higher speed than this without shock to the engine and
fluctuations of air pressure due to agitation of the water piston. The
quantity of heat produced, that is, the degree of temperature reached,
depends entirely upon the conditions in the air itself, as to density,
temperature and moisture, and is entirely independent of speed. We have
seen that it is possible to lose 21.3 per cent. of work when compressing
air to five atmospheres without any cooling arrangements. With the best
compressors of the dry system one-half of this loss is saved by water
jacket absorption, so that we are left with about 11 per cent., which
the slow moving compressor seeks to erase. We are quite safe in saying
that the element of _time alone_ in the stroke of an air compressor
could not possibly effect a saving of more than half of this, or 51/2 per
cent. Now, in order to get this 51/2 per cent. saving, we reduce the speed
of an air-compressing engine from 350 feet per minute to 100 feet per
minute. We must, therefore, in one case have a piston area _three and
one-half_ times that of the other in order to get the _same capacity of
air_, and in doing this we build an engine of enormous proportions with
heavy moving parts. We load it down with a large mass of water, which it
must move back and forth during its work, and thus we produce a
percentage of friction loss alone equal to twice or even three times the
51/2 per cent. heat loss which is responsible for all this expense in
first cost and in maintenance, but which really is not saved after all
unless water injection in the form of spray also forms a part of the

It is obvious that cost of construction and maintenance have much to do
with the commercial value of an air compressor. The hydraulic piston
machine not only costs a great deal more in proportion to the power it
produces, but it costs more to maintain it, and it costs more to run it.
It is not an uncommon thing to hear engineers speak of the hydraulic
piston compressor as the "most economical" machine for the purpose, but
that it is so "expensive" and takes up so much room, and requires such
expensive foundations that, unless persons are "willing to spend so much
money," they had better take the next best thing, a high speed machine.
We hear of "magnificent air-compressing engines, the largest in the
country," and pilgrimages are made to see these artificial wonders when,
not unlike the old pyramids, they represent a pile of inert matter--a
monument to moneyed kings.

The hydraulic piston compressor has one solitary advantage, and that is,
it has no dead spaces. It was conceived at a time when dead spaces were
very serious conditions--were positive specters! Valves and other
mechanism connected with the cylinder of an air compressor were once of
such crude construction that it was impossible to reduce the clearance
spaces to a reasonable point, and, furthermore, the valves were heavy
and so complicated that anything like a high speed would either break
them or wear them out rapidly, or derange them so that leakages would
occur. But we have now reduced inlet and discharge valves and all other
moving parts connected with an air cylinder to a point of extreme
simplicity. Clearance space is in some cases destroyed altogether by
what is, as it were, an elastic air head which is brought into direct
contact with the piston. All this reduces clearance to so small a point
that it has no influence of any consequence. The moving parts are made
extremely simple, even arriving at a point where inlet valves are opened
and closed by their natural inertia. Mr. Sturgeon, of England, has
applied a most ingenious and successful inlet valve, which is opened and
closed by the friction of the air piston rod through the gland. We have,
therefore, reached a point at which high speed is made possible.

Long-stroke air compressors are evidently objectionable on the basis of
greater expense of construction. All the parts must be larger and
heavier. The fly wheels are increased enormously in diameter and weight,
and the strength of bearings must be enlarged in proportion. It is
difficult to equalize power and resistance in air compressors with long
strokes. The speed will be jerky, and when slow, the fly wheel rather
retards than assists in the work of compression. This action tends to
derange the parts and makes large bearings a necessity. The piston in a
long-stroke compressor travels through considerable space before the
pressure reaches a point where the discharge valve opens, and after
reaching that point it has to go on still further against a prolonged
uniform resistance. This makes rotative speed difficult. During the
early part of the stroke, the energy of the steam piston must be stored
up in the moving parts, to be given out when the steam pressure has been
reduced through an early cut-off. With a short stroke and a large
diameter of steam cylinder we are able to get steam economy or early
cut-off and expansion without the complications of compounding.

* * * * *

[Continued from SUPPLEMENT, No. 793, page 12677.]


By G.D. Hiscox.


The natural flow of water in a current is probably one of the oldest and
cheapest of the methods for obtaining power, or the lifting of water
within moderate elevations, for a supply for irrigation and domestic
purposes; and we propose, apart from the current wheel, to treat only of
self-water-raising devices in this chapter.

Water wheels of various forms for this purpose have been used from time
immemorial in Europe, Asia and Egypt, where the record gives examples of
wheels of the noria class from 30 to 90 feet in diameter; the term
_noria_ having been applied to water wheels carrying buckets for raising
water; the Spanish _noria_ having buckets on an endless chain.

Records of a Chinese noria, of 30 feet diameter, made of bamboo, show a
lifting capacity of 300 tons of water per day to a height of 3/4 of the
diameter of the wheel--velocity of current not stated.

For less quantity and greater elevation, these forms of wheel may have
pumps attached to the shaft, by crank, that will give a fair duty for a
high water supply.

For power purposes, as in the plain current wheel, Fig. 23, there are
two principal factors in the problem of power--the velocity of the
current and the area of the buckets or blades.

[Illustration: Fig. 23]

Their efficiency is very low, from 25 to 36 per cent., according to
their lightness of make and form of buckets. A slightly curved plate
iron bucket gives the highest efficiency, thus ( to the current, and an
additional value may also be given by slightly shrouding the ends of the

The relative velocity of the periphery of the wheel to the velocity of
the current should be 50 per cent. with curved blades for best effect.

The most useful and convenient sizes for power purposes are from 10 to
20 feet, and from 2 to 20 feet wide, although, as before stated, there
is scarcely a limit under 100 feet diameter for special purposes.

In designing this class of wheels special attention should be given to
the concentration and increase of the velocity of the current by wing
dams or by the narrowing of shallow streams; always bearing in mind that
any increase in the velocity of the current is economy in increased
power, as well as in the size and cost of a wheel for a given power.

The blades in the smaller size wheels should be 1/4 of the radius in
width, and for the larger sizes up to 20 feet, 1/5 to 1/6 of the radius
in width and spaced equal to from 1/4 to 1/3 of the radius.

They should be completely submerged at the lowest point.

For obtaining the horse power of a current wheel, the formula is

Area of 1 blade x velocity of the current in ft. per sec.

x by the square of difference of velocities of current and wheel
periphery = the horse power; or

A x V 2
------ x (V - v) = h. p.

[TEX: \frac{A \times V}{400} \times (V - v)^2 = h. p.]

in which A equals the area of blade in square feet, V and v velocities
of current and wheel periphery respectively, in feet per second. Thus,
for example, a wheel 10 feet in diameter with blades 6 feet long and 1
foot in width, running in a stream of 5 feet per second--assuming the
wheel to be giving as much power as will reduce its velocity to one half
that of the stream--the figures will be

6' x 5' 2
------- x 2.5 = 0.468

[TEX: \frac{6' \times 5'}{400} \times 2.5^2 = 0.468]

horse power of the wheel.

The total power of the stream due to the area of the blade equals the

Square of the velocity of the stream
------------------------------------ x
Twice gravity (64.33)

volume of water in cubic feet per second x 62.5 (weight of 1 C') = the
value or gross effect in pounds falling 1 foot per second. This sum
divided by 550 = horse power. Thus, as per last example,

------ x 30 x 62.5
---------------------- = 1.32 the horse power of the current

[TEX: \frac{\frac{5^2}{64.33} \times 30 \times 62.5}{550} = 1.32 \text{
the horse power of the current}]

due to the area of the blades of the water wheel.

For the efficiency of this class of wheel, with slightly curved and thin
blades, divide the horse power of the wheel by the horse power of the
current area, equals the percentage of efficiency.

As in the last case,

0.468 / 1.32 = 0.351/2

per cent. efficiency of the water wheel.

With higher velocities of stream and wheel the efficiency will be from 2
to 3 per cent. less, although the horse power will increase nearly with
the increase in velocity of the current.

For details of application of various forms of current wheels for power
purposes see illustrated description Yagn's and Roman's floating motors

A very good example of a floating motor of the propeller class is
Nossian's fluviatile motor, illustrated and described in SCIENTIFIC

[Illustration: Fig. 24.]

Fig. 24 represents a very complete floating motor, in which the floats
are wedge shaped at the stem, for the purpose of increasing the current
between them, the wheel being an ordinary current wheel, as shown in
Fig. 23, with a curved shield or gate in front, which can be moved
around the periphery of the wheel for the purpose of regulating its
speed or stopping its motion by cutting off the stream from the buckets.

The float, rising and falling with the stream, is held in position by a
braced frame swinging on anchorages within the mill on shore, and
parallel with a swiveled shaft.

Tide wheels and tidal current wheels have been in use for more than 800
years, and were largely in use in Europe and the United States during
the first half of the present century. No less than three were running
in the immediate vicinity of New York, in 1840, for milling purposes.

Their day seems to be past, except in some special localities. We will
also pass them, and illustrate some of the


The tympanum derives its name from its similarity to a drum as made by
the Romans, but its origin was Egyptian. It is a current wheel with
frame like Fig. 23, to the outside of which a set of chambers or tubes
are fixed, radiating spirally, so as to lead the water to the shaft as
the wheel revolves, as shown in Fig. 25. It has a lift of a little less
than half its diameter, and answers an excellent purpose for the
irrigation of rice and cranberry fields, or on streams running through
low lands in arid districts. It is still one of the Nile irrigating

[Illustration: Fig. 25]

The building of these wheels is within the scope of the carpenter and
the tinsmith. A short wooden shaft made square or octagonal, as
convenient, with gudgeons in the ends and arms of wood bolted across
each of the sides of the shaft, or as shown in the cut, will form a
frame work upon which a rim may be fastened, to which the blades and
tubular buckets can be attached.

The directions in regard to the current wheel, Fig. 23, may be followed
as to number and form of blades, which must be made in length and width
proportional to the velocity of the stream and the quantity of water to
be lifted by each tubular arm. The tubes may be made of galvanized sheet
iron and attached to the outside of the wheel, as shown in Fig. 25.


This is a simple current wheel with pot buckets, rigid or swinging,
arranged on the rim of the wheel, to carry up and discharge the water
nearly at the top of the wheel, and through the long ages that it has
been in use for irrigation, village water supply, and even for private
establishments, has assumed a variety of forms in detail of construction
ranging from the bamboo wheels of the Chinese to the light iron wheels
of modern construction.

We illustrate the most simple of these forms in Figs. 26 and 27, in
which the first is a series of boxes or chambers in the rim of the wheel
with side openings in the forward part of the box as the wheel revolves,
and a lip extending from the inner edge of the opening to direct the
outflow into the trough.

[Illustration: Fig. 26.]

Another form, Fig. 27, is arranged with swing buckets or pots, pivoted
just above their centers, and with the catch trough so fixed as to tip


Back to Full Books