that the apprehensions generally entertained with regard to fans of large size are unfounded, and also that compared with such fans any form of piston or constant-volume ventilator loses its one apparent advantage over centrifugal ventilators, viz., the production of a better vacuum. The occasion for this fan arose from the Company finding it necessary to double the output from one of its mines in order to work the upper and lower seams at the same time. This mine had three shafts, one of which was fitted with a furnace, and another with a Guibal fan 23 feet in diameter. The Company decided on sinking a fourth shaft, 13.1 feet in internal diameter, but until this could be completed (a time of more than two years), it was necessary to provide apparatus for drawing an additional volume of 35 to 40 cubic mètres (46 to 52 cubic yards) per second through the existing workings. A Lemielle ventilator, 23 feet high and 16 feet in diameter, was tried for this purpose; but experiments showed that, at an extra speed of 25 revolutions, the quantity discharged was only 25 384 cubic mètres per second, with a vacuum represented by 0.143 mètre of water (5.6 inches); and that in spite of all possible care in construction and maintenance the ordinary speed of working could not be taken at more than 18 revolutions, which gave a volume of 19.304 cubic mètres, with a vacuum of 0.098 mètre (3.9 inches). The addition of a second Lemielle ventilator would of course have given the discharge required; but this was decided against, because experience proved that with the Lemielle ventilator frequent stoppages cannot be avoided from accidents occurring either to the vanes themselves or to the rods which actuate them. It was also to be remembered that in two years a large new shaft would be available for the return of air; and the main reason for requiring at the moment a very high vacuum was, that the existing return shafts were of small section, and a great difference of pressure was therefore necessary to force a sufficient quantity of air through them. This was shown by experiments made in the mine itself. The result of the new shaft would thus be to improve the "temperament of the mine (viz., the ratio of the square of the discharge to the corresponding height of water column). The question which form of ventilator would answer best with this improved temperament had been solved at a neighbouring colliery, where a Lemielle ventilator, 16 feet high, 23 feet in diameter, and a Guibal fan, 6.6 feet long, 29 feet in diameter, were in connection with the same shaft; these were both worked with a connection between the shaft and the surface alternately closed and open, so as to give a greater resistance in the first case than in the second. In both cases the Lemielle was worked at 12.77 revolutions, and the Guibal at 59 revolutions per minute. With both ventilators the discharge rose in the second case; but in the Lemielle it was only from 27.041 cubic mètres to 28.036 cubic mètres, while with the Guibal it was from 27·090 cubic mètres to 30 030 cubic mètres. This advantage on the side of centrifugal ventilators, combined with their less liability to accident and

greater facility of maintenance, decided the Company to give them the preference over those of constant volume.

The principle of the fan being thus adopted, and the Guibal selected as the best existing type, it remained to consider whether a small fan at high speed or a larger fan at more moderate speed would best answer the purpose. Two methods have been used in the district for working fans up to a high speed. The first is by belts and pulleys of different dimensions, which acts well for volumes of 15 to 20 cubic mètres of air; but beyond this the great size required for the pulleys and the loss of power in the belts render the method inapplicable. The second is by toothed wheels and pinions driven by engines provided with variable expansion gear; but with this system it has been found impossible to prevent constant shocks which endanger the teeth; these latter also wear unequally, those which act at the time of full pressure lasting a much shorter time than the others. From these considerations it was resolved to have a fan of large diameter and moderate speed. At this time the Brotherhood three-cylinder engine was attracting much attention; and as its simplicity and speed of working appeared to promise favourable results, one was fitted to a Guibal fan, 23 feet in diameter, which was available for the pose. The engine was worked for nearly a year, but finally abandoned, as it was found impossible, with any of the means of lubrication employed, to prevent heating and seizing of the moving parts in the cylinders. A horizontal "Meyer" expansion engine was ultimately put down for the new fan, having a diameter of 0.68 mètre (26.6 inches), and a stroke of 0.85 mètre (33.5 inches). Accidents with large fans have generally been due either to vibrations of the shaft or to defective keying of the arm centres. In this case a forged shaft of 0.32 mètre (13 inches) was used, which proved perfectly rigid; and besides the ordinary keying, a cast-iron boss was placed on the shaft between the centres, and secured to it by a spiral row of powerful binding screws. In the construction of the fan itself, the masonry roof, and the wooden casing, every precaution was taken to guard against the possibility

of accident.

pur

The final specification was that the fan should be 12 mètres in diameter (39 feet 5 inches), 23 mètres wide (8 feet 2 inches), and should be capable of drawing from the mine 40 cubic mètres (52) cubic yards) of air per second, with a vacuum of 200 millimètres (77 inches) of water, at a speed of 80 revolutions per minute, and with a steam pressure of 3 atmospheres. Experiments were made by two independent engineers to ascertain whether the fan, as erected, fulfilled these conditions. These were made with the mine in its ordinary state of working. The velocity of the air was measured in the passage leading from the shaft to the fan (the section of which was 57.8 square feet), by means of a Biram's anemometer, carefully verified beforehand. The fan was to have been tried at three different speeds, 50, 60, and 80 revolutions per minute; but the steam supplied was insufficient to produce more than a speed

varying from 76 to 79 revolutions per minute, the mean being 77.5. The results are given below:

It will be seen that the fan had thus more than fulfilled the conditions as to quantity of air extracted, while its construction and working were reported as being in every respect satisfactory. Indicator diagrams of the engine were taken at the same time, with a view of ascertaining the useful effect, but it was found that at the higher speed these were a failure. For the speeds of 50 and 60 revolutions they gave a useful effect (or ratio between the power developed in the air passage and in the cylinder) of 0.5385 and 0.5551 respectively. Observations were also made of the air-pressure within the chimney of the fan, by which the air escapes. For this purpose the tube of a manometer was passed through the brickwork, and gradually pushed farther and farther from the wall, so as to measure the depression of the water column for each part of the chimney. This was done at three places, the heights of which above the actual outlet of the fan were 4 inches, 11.2 feet, and 27.3 feet respectively. With a depression in the fan of 3.5 inches, the depression at the lowest place varied from 0.4 inch close to the wall to 1.85 inch at the centre; giving a mean depression of 0.93 inch, or a ratio of 0.26 between the depression in the chimney and in the fan. At the second place, the variation was from 0.04 inch at the wall to 0.48 inch at 4.3 feet from it. At the highest place, the variation was from 0.04 inch at the wall to 0.22 inch at 2 feet from it. Similar results were obtained with depressions in the fan of 4.6 inches and 5.9 inches. On the whole the experiments showed that the velocity throughout the area of the chimney was tolerably uniform, and that consequently there could be no fear of any re-entering currents.

W. R. B.

On Ventilating Shafts in Coal Stores. By H. HENNING.

(Journal für Gasbeleuchtung, vol. xx., pp. 518-520.)

Messrs. Abel and Percy, in the report upon ventilation and spontaneous combustion in coal,1 held that ventilation on board ship

1

"It does not appear practicable, therefore, to apply ventilation with any prospect of guarding against the accumulation of heat in some portion of a cargo of

was impossible, and that in stores ventilating shafts were not only useless, but dangerous. The following facts seem to point to a different conclusion.

The Author had to erect (at Danzig) a coal-shed of considerable size. On two previous occasions extensive fires occurred in which 1,000 tons of Silesian coal, and in another instance 2,800 tons of English coal were partly reduced to coke and partly consumed. Observations made at the time showed that the heated air could leap over distances of 10, 20, and 25 feet without any perceptible cause. These facts induced the Author to introduce ventilating shafts in the new sheds, in spite of the various opinions against their application. Two thousand tons of very damp English coal were put into the new shed. The ventilating shafts, in the absence of any data as to their distance, were placed at random, 25 feet apart in one direction and 12 feet in the other, and communicated at the bottom with underground channels.

The amount of moisture in the coal was about 2 per cent. ; and about a fortnight later the whole shed was filled with steam, and the air in the ventilators had risen to 95° and 113° Fahr. The shafts were provided with valves, so as either to admit or exclude the external air. These were immediately opened, upon which there was so rapid an evolution of steam, that the next day the evaporation had nearly ceased, and after eight days no more vapour was visible. The coals were then removed, and it was found, that in the direction of the 12-feet distance they were perfectly cold, while in the 25-feet direction the middle portion was warm for a length of about 12 feet. This was a clear proof that the distance of 25 feet was too great. The height to which the coal was stored was 20 feet.

From these facts the Author concludes that ventilating shafts, placed at distances equal to one-half the height to which the coal is stacked, will under all circumstances have a good effect; and if a thermometer is inserted into the air shafts, a sure test of the condition of the coal is constantly at hand.

J. G. H.

On Turbines with Partial Distribution. By PROF. RICHELMY.

(L' Ingegneria civile e le Arti industriali, vol. ii., pp. 64, 115, 135, 155.)

An essential difference ought to be maintained between the arrangement of turbines with partial distribution, and that of turbines with total distribution. For the latter, the space com

coal. Indeed, a period would be reached when the development of heat would be most seriously promoted by ventilation, some time before actual ignition demanded the total exclusion of fresh air."-Paper by F. A. Abel, Esq., F.R.S., and J. Percy, Esq., M.D., F.R.S., in the Report on the Spontaneous Combustion of Coal in Ships, 1876, p. xxvi.

prised between the supply chamber and the motor should generally have no contact with the atmosphere; while, with those for partial distribution, the atmosphere should always penetrate into that space. If, in fact, this do not take place, the passages constituting the motor, comprised between vane and vane, will be subject to a rapid and injurious variation in the pressure against the upper face. When the passage is in front of one mouth of the supply chamber, that pressure will correspond to the head in the supply chamber when the passage passes in front of the closed part of the supply chamber, the pressure will rapidly cease.

Presupposing the free circulation of the air, between the supply chamber and the motor, the theory becomes very simple. The water in the supply chamber moves as in a closed pipe, the two extremities of which are subjected to atmospheric pressure only. The head, and the sum of the orifices are known; whence it becomes easy to calculate the theoretical discharge. Some loss may occur in the head by friction and by reason of the passages between successive sections being appreciably different in area. Many experiments made with a view to discover the actual amount of all these losses have induced the Author to estimate them at between five hundredths and eight hundredths of the head, according to circumstances. Nevertheless, the effective discharge being somewhere between ninety-two hundredths and ninety-five hundredths of the theoretical, and the efflux taking place from the orifices of the supply chamber without contraction, the diminution ought necessarily to be attributed to a lesser velocity of efflux, due no longer to the head, but only to this multiplied by (0.92)2 or (0.95)2, or by a number intermediate between these two. The head upon the discharging orifices from the supply chamber, for many wheels, is almost equal to the fall; whence, for all these turbines, from 10 to 15 per cent. of the energy of the stream of water is lost by friction, by bends in the pipes and passages, and by the sudden changes of section which precede the outflow from the supply chamber. Then, when, having left this first organ of the apparatus, the water is ready to enter into the motor, it is not difficult to estimate the relative velocity with which it proceeds to do so. Resolving the absolute velocity of the efflux into two components, the first of which is the velocity of rotation of the parts of the motor where the water is introduced, the second component is the relative velocity sought. The water, being left with this velocity inside the spaces comprised between the one and the other of two consecutive vanes, will be able to preserve it, or will be obliged to change it. This latter is the case which under a certain aspect may be said to be more frequent. Then the relative velocity, which the water must take, is determined by means of the law of continuity, and has a direction which is fixed by the inclination of the vane to the circumference of the wheel. There are, then, two different relative velocities, the second of which is altogether independent of the intensity of the rotary velocity of the motor; but the first of which depends on it: whence, in order