The coal was in a finely divided state, being in grains from 2 to 4 millimetres in diameter; and, after repeated exhaustions, the apparatus was allowed to stand for 3 or 4 days, before the heating with a water-bath was commenced. In this way, as Dr. Broockmann states, he obtained a minimum quantity of gas, and he claims that the gas so obtained is truly representative of the "enclosed gases," and is not contaminated by air adhering to the coal, or by those gases produced by the action of the air on the coal itself.

Experimenting with the Ryhope coal, Dr. Broockmann obtained 70 cubic centimetres of gas, which consisted of 97 per cent. of combustible gas, and 3 per cent. of carbon dioxide; that is, from 100 grammes of coal he obtained 679 cubic centimetres of combustible gas, and 21 cubic centimetres of carbon dioxide. Whereas, as will be seen by reference to the paper by the late Mr. McConnell and the author, they obtained from the same weight of coal 6107 cubic centimetres of combustible gas, 42 cubic centimetres of carbon dioxide, 140 cubic centimetres of oxygen and 5553 cubic centimetres of nitrogen, making a total of 122:2 cubic centimetres.

As the writer's object in studying the gases enclosed in the Ryhope coal was to obtain information which would throw some light on the nature of the combustible gases that he had previously found in the dust from the screening of this coal, the coal was not taken in small fragments, but in such fragments as would permit to some extent of the partial separation (by reason of the differences in rates of effusion) of the denser from the lighter hydrocarbons. Thus, a rough fractionation would be effected, and in the results of the analysis of the several fractions the presence of different paraffin hydrocarbons would be more clearly indicated.

To return to the use of indiarubber-stoppers in investigations of this kind: it is certainly true that in the experiments on the gases enclosed in coal-dust, the author employed indiarubberstoppers, but with the precaution that these stoppers were covered by a layer of Faraday cement. In this connection it may be of interest to record the result of a recent experiment on the manner in which vessels closed in this way will maintain a vacuum over a lengthened period. A flask, of about 300 cubic centimetres capacity, was closed by a tightly fitting indiarubber-stopper * Trans. N.E. Inst., 1888, vol. xxxvii., pages 245 to 256.

through a hole in which was inserted a glass-tube by which the flask was sealed on to a Geissler pump. The cork was carefully covered with Faraday cement, and then the flask was exhausted and closed off from the pump. After standing some two months, the flask was again exhausted, and the gas obtained collected over mercury. A single bubble of gas was thus obtained, the volume of which proved to be approximately 0.2 cubic centimetre. As this experiment was made under conditions practically similar to those described by the author in his earlier papers,* it may be concluded that, in these experiments, the disadvantages arising from the use of indiarubber-stoppers had been satisfactorily

overcome.

A second point in Dr. Broockmann's criticism is the assumption that the author had overlooked the results obtained by the Austrian Fire-damp Commission in the examination of certain varieties of coal-dust. This is all the more surprising, since in the paper in the Transactions (to which, as had been already mentioned, Dr. Broockmann referred), and also in the account of the lecture given by the writer to the members of the National Association of Colliery Managers at Nottingham,† the work of the Austrian Fire-damp Commission is specially mentioned. That the writer should have regarded the experiments of the Austrian Fire-damp Commission as confirmatory of his own observations cannot appear strange or remarkable, as the author's first paper on the subject was read before this Institute in August, 1888, while the investigation by the Austrian Fire-damp Commission of the coal-dust question was made during the years 1889 to 1891. Further, the conclusion arrived at by the Commission, and cited by Dr. Broockmann, "that the content of dense, easily inflammable hydrocarbon gases increases both the sensitiveness and the dangerous character of a coal-dust," appeared in the Final Report of this Commission published in 1891. This will suffice to show the independence of the two sets of observations.

Another observation of the late Mr. McConnell and the author, selected for special criticism by Dr. Broockmann, was the composition of the gas obtained by heating the coal at 1840 Cent. As Dr. Broockmann pointed out, this gas was remarkable for the large percentage of oxygen given in the analysis; but he, unfortunately,

* Loc. cit.

+ Colliery Manager and Journal of Mining Engineering, 1895, vol. xi., pages 30 to 32.

had not troubled to state exactly the conditions under which the gas was produced, nor the volume of gas obtained under these conditions from 100 grammes of coal. Had this been done, the reader of Dr. Broockmann's paper would, in the first place, have been able to appreciate aright the bearing, on the point under discussion, of the experiments made by Dr. Broockmann to demonstrate the nature of the changes produced in air when it is heated in closed tubes with coal to a temperature of 160° Cent. In the second place, the reader would have been compelled to admire the ingenuity of Dr. Broockmann, who credits the writer with the discovery of a new method of preparing oxygen, and this because from 100 grammes of coal he had obtained 21 cubic centimetres of oxygen, or barely 3 parts by weight of oxygen from 100,000 parts by weight of coal.

In the paper read before this Institute in February, 1894, will be found a description of the experiment from which this result had been isolated.* And there it will be noted that a definite weight of coal was introduced into a tube, sealed off at one end and at the other end sealed on to an air-pump. After exhausting the air from the tube, the coal was heated for a certain period by passing steam through a jacket surrounding the tube which held the coal. The gas so produced was drawn off; then, after the coal had ceased to yield gas at this temperature, it was heated by passing the vapour of amyl alcohol through the jacket. When the coal no longer yielded any gas at this higher temperature, it was heated for some hours to a still higher temperature, by passing the vapour from boiling aniline through the jacket; and thus the further fraction of gas was obtained, which in this instance formed from 4 to 5 per cent. of the total volume of gas extracted from the coal.

A repetition of this experiment became necessary, in the light of the importance attached to the composition of the gas obtained by this extraction at 1840 Cent., and the difficulty in finding an explanation of the proportion of oxygen which it was found to contain; a difficulty made the greater by Dr. Broockmann's statement that the gas was obtained in experimenting with fine coal, whereas it was produced from coa! in pieces, and pieces certainly much larger than those he employed in his investigation. It is not inconceivable that, under the conditions of the experiment, the heating of the coal would cause a mechanical breaking-up of Trans. Inst. M.E., 1894, vol. vii., page 35.

the pieces of coal, and thus favour the release of oxygen and other gases not already removed in the previous extractions. As to the proportion of oxygen to the other constituents, while the statement in volume per cent. may appear misleading, it should be stated that a careful scrutiny of the actual analytical data, from which these percentages were computed, had revealed an error in calculation. But, even when this correction is made, the proportion of oxygen to nitrogen is larger than the relative amount of these gases in the air. As to the contention that this observation is contradictory of all our knowledge of the mutual behaviour of air and coal at such temperatures, and its refutation by the experiments recorded by Dr. Broockmann, it will be sufficient to point out that the chemical relations of many bodies are altered by conditions of pressure, and that conclusions drawn from the experiments made under increased pressure are not applicable to explain the results of experiments made under greatly reduced pressures. The impossibility of explaining the presence of the constituent gases of the atmosphere in this gas in the manner suggested by Dr. Broockmann has already been referred to, and that oxygen and nitrogen are found in the gases extracted from coal under these conditions is shown by the results of experiments Nos. IV. and V. described below.

Turning now to the results obtained in the recent investigation of the gases enclosed in the Hutton-seam coal, it will be sufficient to point out that the method of extracting the gases is that employed by Messrs. McConnell and Bedson, described in the paper already referred to.* In experiments Nos. I. to IV. the pump employed, to which the tube containing the coal was sealed on, was one of the Geissler type; whereas in No. V. the pump used was a form devised by Prof. Töpler, without taps. After establishing a vacuum in the pump and tube containing the coal, the apparatus was allowed to stand at the ordinary temperature for some days, and then the gas produced in the interval was drawn off. The volume of the fraction of gas so obtained was measured and afterwards analysed. In order to obtain, if possible, a clearer idea of the nature of the combustible constituents, and to demonstrate as far as practicable the composition of these combustible gases, the coal was submitted to a prolonged * Trans. Inst. M.E., 1894, vol. vii., page 27.

extraction at the ordinary temperature, before heating with steam was resorted to, and in this way a series of fractions was obtained, each being separately analysed.

The writer's engagements have in many instances determined the lengthy duration of some of these operations: but some compensation for this delay may possibly be found in the information supplied by the results of the analyses of the several fractions of gases obtained from the coal under these circumstances.

In experiment No. IV., instead of removing the air from the tube holding the coal in the ordinary way, the coal was filled into a tube, in which it was held in position by plugs of glass-wool; the tube was next drawn out at both ends, and on to these narrow glass-tubes were sealed. By one of these narrow tubes, the tube was sealed on to the air-pump; while to the other, which was over 30 inches in length, a movable vessel containing mercury was attached by a stout indiarubber-tube. The wider section of the tube was surrounded by a second which formed a jacket for the passage of steam, etc. To expel the air from the tube, the tap of the pump was opened, and by raising the vessel containing mercury, the whole was filled with mercury up to the tap of the pump, which was then closed. The movable vessel was next lowered, the end of the glass-tube to which this vessel was attached stood in a vessel filled with mercury, and on removal of the indiarubber-tube there was thus established a rough barometric column, with the coal standing in a partial vacuum. The height of the mercury was found to be approximately 710 millimetres. After standing overnight, the exhaustion was completed in the ordinary way; and after remaining in this condition for 9 days, the gas given off in the interval was drawn off. Thus it will be seen that the coal was sealed off from the air, not by closing the tube by fusion of the glass, but by a barometric column of mercury. This method of exhaustion was practised in experiment No. V. and afforded a quicker means of removing the air from the tube than by the ordinary method of repeated pumpings, and one which consequently yielded a larger amount of combustible gas per given weight of coal.

Below are given in tabulated form the results of the measurements and analyses of the several fractions of the gases extracted from the coal in the manner indicated. The results of experiments Nos. I. and II. are not stated in detail, as the analyses of

VOL. XXIV-1902-1903.

3