differences of electrification due either to irregularities in the drops or to differences of situation, and is at first difficult of acceptance in view of the efficiency of such very feeble electric forces. Fortunately I am able to bring forward additional evidence bearing upon this point.

When two horizontal jets issue from neighbouring holes in a thin plate, they come into collision for a reason that I need not now stop to explain, and after contact they frequently rebound from one another without amalgamation. This observation, which I suppose must have been made before, allowed me to investigate the effect of a passage of electricity across two contiguous water surfaces. The jets that I employed were of about inch in diameter, and issued under a moderate pressure (5 or 6 inches) from a large stoneware vessel. Below the place of rebound, but above that of resolution into drops, was placed a piece of insulated tin plate in connexion with a length of gutta-percha-covered wire. The source of electricity was a very feebly excited electrophorous, whose cover was brought into contact with the free end of the insulated wire. When both jets played upon the tin plate, the contact of the electrified cover had no effect in determining the union, but when only one jet washed the plate, union instantly followed the communication of electricity, and this notwithstanding that the jets were already in communication through the vessel. The quantity of electricity required is so small that the cover would act three or even four times without being re-charged, although no precautions were taken to insulate the reservoir.

In subsequent experiments the colliding jets, about 1 inch in diameter, issued horizontally from similar glass nozzles, formed by drawing out a piece of glass tubing and dividing it with a file at the narrowest part. One jet was supplied from the tap, and the other from the stoneware bottle placed upon an insulating stool. The sensitiveness to electricity was extraordinary. A piece of rubbed guttapercha brought near the insulated bottle at once determined the coalescence of the jets. The influencing body being held still, it was possible to cause the jets again to rebound from one another, and then a small motion of the influencing body to or from the bottle again induced coalescence, but a lateral motion without effect. If an insulated wire be in connexion with the contents of the bottle, similar effects are produced when the electrified body is moved in the neighbourhood of the free end of the wire. With care it is possible to bring the electrified body into the neighbourhood of the free end of the wire so slowly that no effect is produced; a sudden movement of withdrawal will then usually determine the coalescence.

Hitherto statical electricity has been spoken of; but the electromotive force of even a single Grove cell is sufficient to produce these phenomena, though not with the same certainty. For this purpose one pole is connected through a contact key with the interior of the

stoneware bottle, the other pole being to earth. If the fingers be slightly moistened, the body may be thrown into the circuit, apparently without diminution of effect. This perhaps ought not to surprise us, as in any case the electricity has to traverse several inches of a fine column of water. On the other hand, it appeared that most

of the electromotive force of the Grove cell was necessary.

66

Further experiment showed that even the discharge of a condenser charged by a single Grove cell was sufficient to determine coalescence. Two condensers were used successively; one belonging to an inductorium by Ladd, the other made by Elliott Brothers, and marked CapacityFarad." Sometimes even the "residual charge " sufficed. It must be understood that coalescence of the jets would sometimes occur in a capricious manner, without the action of electricity or other apparent cause. I have reason to believe that some, at any rate, of these irregularities depended upon a want of cleanness in the water. The addition to the water of a very small quantity of soap makes the rebound of the jets impossible.

The last observation led me to examine the behaviour of a fine vertical jet of slightly soapy water; and I found, as I had expected, that no scattering took place. Under these circumstances the approach of a moderately electrified body is without effect, but a more powerful influence scatters the drop as usual. The apparent coherence of a jet of water when the orifice is oiled was observed by Fuchs, and appears to have been always attributed to a diminution of adhesion between the jet and the walls of the orifice.

Some further details on this subject, and other investigations respecting the phenomena of jets, are reserved for another communication, which I hope soon to be able to present to the Royal Society; but I cannot close without indicating the probable application to meteorology of the facts already mentioned. It is obvious that the formation of rain must depend very materially upon the consequences of encounters between cloud particles. If encounters do not lead to contacts, or if contacts result in rebounds, the particles remain of the same size as before; but, if the issue be coalescence, the bigger drops must rapidly increase in size and be precipitated as rain. Now, from what has appeared above we have every reason to suppose that the results of an encounter will be different according to the electrical condition of the particles, and we may thus anticipate an explanation of the remarkable but hitherto mysterious connexion between rain and electrical manifestations.

VOL. XXVIII.

2 H

II. "On the Influence of Coal-dust in Colliery Explosions." No. 2. By W. GALLOWAY. Communicated by ROBERT H. SCOTT, F.R.S., Secretary to the Council of the Meteorological Office. Received February 27, 1879.

In the former communication on this subject, which I had the honour of submitting to the Fellows ("Proc. Roy. Soc.," vol. xxiv, p. 354), some experiments were described which showed that a mixture of air and coal-dust of a certain known chemical composition was not inflammable at ordinary pressure and temperature; and that, when 0.892 per cent. of fire-damp (by volume), or a greater proportion, was added to the same mixture, it became inflammable and burned freely with a red smoky flame. The general conclusion to which the second result pointed was also stated in the same place to be, that, an explosion originated in any way whatever in a dry and dusty mine, may extend itself to remote parts of the workings, where the presence of fire-damp was quite unsuspected.

The wetness or dryness of the workings of a mine depends, other things being equal, on the temperature of the strata in which they are situated: for it is obvious that if, on the one hand, the temperature of the mine is lower than the dew-point of the air at the surface, the ventilating current will deposit moisture as it becomes cooled in passing through the workings; and if, on the other hand, the temperature of the mine is higher than the dew-point of the air at the surface, the ventilating current will absorb moisture and tend to produce a state of dryness. It is well known, however, that the temperature of the strata in the Coal Measures of this country increases at the rate of about 1° F. for every 60 feet of additional depth below the surface, and, therefore, from what precedes, it is evident that the comparative wetness or dryness of a mine depends on its depth.

As far as my own observations are concerned, I have found that coal mines, shallower than 400 feet, are damp or wet, and those deeper than 700 feet are dry and dusty: between these two points, also, there appears to be a kind of debateable ground in which wetness or dryness depends, for the time being, on the coldness or warmness of the air entering the mine at the surface.

In all dry coal mines the coal-dust lying on the floor of the roadways rises in clouds and fills the air when it is disturbed by the passage of men, horses, small waggons, &c.; a sudden puff of air, therefore, such as that produced by a local explosion of fire-damp, or by a shot blowing out its tamping, must necessarily produce the same effect in a greater or less degree according to its intensity. The mixture of coal-dust and air, formed by the action of either the fire

damp explosion or the blown-out shot, will be inflammable if it contain any larger proportion of fire-damp than 0.892 per cent., and the flame of the original explosion will pass on through it, extending the area of the disturbance as far as the same conditions exist, or, it may be, to the utmost limits of the workings. If it contain more than 0-892 per cent. of fire-damp, it will be more and more explosive, according as the proportion of fire-damp is greater, until a maximum point is reached, beyond which its explosiveness will begin again to decline. If, lastly, it contain less than 0.892 per cent. of fire-damp, or even if it consist only of coal-dust and pure air, it will still be so nearly inflammable that it will probably become so when it undergoes the compression and consequent heating which the occurrence of an explosion in one part of a confined space must necessarily produce throughout the remainder of the same space. It is probable, moreover, that some kinds of coal-dust require less fire-damp than others to render their mixture with air inflammable; and it is conceivable that still other kinds may form inflammable mixtures with pure air.

I have partially investigated the relation between the proportions of air, coal-dust, and gas* required to insure inflammation or explosion on the application of a light; but as the series of experiments is not yet complete, I propose to reserve their description for some future opportunity. I may mention, however, that in the apparatus which I have hitherto employed, the proportion of coal-dust which gave the best results was much larger than might at first sight be thought necessary, namely, about one ounce of dust to a cubic foot of air for all mixtures of gas and air, ranging between one of gas and twenty of air, and one of gas and forty of air. Also, in one of the experiments with the return air of a mine, which I propose to describe in this place, the air requires to be literally black with dust before it will ignite. It is, therefore, obvious that the particles which are floating about in the air of a dry mine, in its normal state, cannot render it inflammable; and it is probable that only the sweeping action of a gust of wind, like a squall, passing along the galleries, can raise a sufficient quantity to do so.

Some of the colliery explosions which have occurred during the last two years are amongst the most disastrous on record, and the attempts that have been made to explain them are of the usual unsatisfactory character. The assumption, without a vestige of proof that fire-damp has suddenly burst from the strata, is still maintained even in cases in which the flame is seen to have ramified into the extremity of every cul-de-sac and extended to the opposite boundaries of the workings. The very token whereby the ubiquity of the flame is made manifest, is the so-called charring of the timber, * As these experiments were only preliminary ones made chiefly for the purpose of testing the apparatus, common lighting gas was employed in them.

coal, and rubbish; and this, generally in the case of the timber, and always in the case of the coal and rubbish, consists of a coating of coked coal-dust adhering to them superficially, and testifying unmistakably by its presence that coal-dust has actually been playing the part which is claimed for it by myself and others.

Following are a few of the details that have become known regarding the most recent explosions of importance :

Pemberton (11th October, 1877). 36 men killed. Depth 1,005 feet. At page 333 of the "Reports of the Inspectors of Mines," it is said:"The Pemberton Colliery had been held up as a model of engineering, and seemed to be the last place at which a disaster of this kind was likely to happen." At page 332 of the same volume, the following gratuitous explanation of the explosion is given :-"The effect of a shot blowing out, and which appears to have occurred, would be to exhaust the face and sides of Rutter's place and Price's place, and this additional fire-damp rushing out into an atmosphere already heavily charged, would bring the air in this particular district up to the explosive point."

Blantyre (22nd October, 1877). 207 men killed. Depth of the workings from 800 to 900 feet. The seam is not very gaseous and the mine was supposed to be well ventilated. It was impossible to say where the explosion began. At page 7 of the official report it is said:"The explosion extended throughout miles of the workings and was of the most violent kind. The gas in a large portion of the workings had apparently been mixed to a highly explosive state. The noise at the top of No. 2 shaft is described as having been like a shot in a sinking pit, and a great volume of smoke and dust came to the top. On the top of No. 3 shaft the noise was like the bursting of a steam pipe, or shot in a sinking pit, and was as quickly over, flame coming out of the shaft mouth. Flame seems to have extended through nearly all the working places." Again, at page 11 :-" The mine being dry and dusty, and the dust being mixed with highly inflammable splint coal, would help to spread the flame and give force to the explosion." Lastly, at page 206 of the notes of evidence, a witness says:-"I desire to make a suggestion. On one occasion Mr. Watson (the manager) told me that the mine, being a dry one, like a desert, the coal-dust would aggravate an explosion."*

Unity Brook (12th March, 1878). 43 men killed. Depth of the workings 792 feet. The workings were examined, and found to be safe, half an hour before the explosion. The mine was dry and dusty.

It is encouraging to observe that the agency of coal-dust has thus been recognized by some persons connected with mining, although it appears to have entirely escaped the notice of every one except Faraday and Lyell, and some of the French mining engineers, until after the appearance of my first paper on this subject in