July 3, 1914 THE COLLIERY GUARDIAN. 23 Table A. Month. Accidents. Difference of temperature from mean temperature. Relative quantity of air circulating (A.) Weight of vapour per I cu. ft. air at mean , temperature and humidity for month. Grrs. Weight of water re- quired to bring air to i saturation at mean temperature 60'8 (B). Grs. Relative quan- tity of water absorbed,or relative drying effect (A x B.) January 11 + 8’4 2’90 5’3 + 0’6 1’7 February ... 4 + 7’9 2’82 5’6 0’3 0’8 March 6 + 4*8 2’20 5’1 0’8 1’8 April 11 - 0-6 0’78 . 4’1 1’8 1’4 May 9 - 6-3 2’51 3’5 2’4 6’0 June 10 -11-9 3’45 2’8 3’1 10’7 July 14 -12-5 3’54 2’5 3’4 12’0 August 10 - 6-7 2’59 2’9 3’0 7’8 September ... 6 - 0-2 0’45 3’1 2’8 1’3 October 11 + 3'3 1’82 3’8 2’1 3’8 November 4 + 5-8 2’41 4’5 1’4 3’4 December 9 + 7’8 2’79 5’2 0’7 1’9 roof depends not only on the temperature and percent- age of water vapour in the air, but also on the quantity of air circulating, and that is affected by the difference of temperature within and without the mine. As mean annual temperatures have varied little, temperature has only slightly influenced the relative drying effects of the air in different years. Changes in that effect have depended almost wholly on humidity. The relation of high annual accident rates to low mean humidity has already been observed. Though temperature varied little from year to year, it varied considerably from month to month, and, owing to their double effect, such variations influenced the drying or moistening of the roof more than the Variations in humidity. It is impossible to calculate accurately the effect of the air, but some further progress with our investigation may be made by assuming as , crude approximations that the air, when in the mine, reaches the mean temperature (60'8 degs. Fahr.) and becomes saturated and that (as nearly all colliery ventilation in the district is natural), the quantity of air circulating varies as the square root of the difference of mine and surface temperatures. On those assumptions we arrive at the figures shown in Table A. The results are plotted as a curve in diagram 2. It will be seen that the last column shows a drying action in each month, with maximum in July, which is the month in which accidents reach their maximum. Making a quarterly comparison, we have :— Deaths. Accidents. Average relative drying. January to March ... ... 12 . .. 21 ... 1’42 April to June ... 18 . .. 30 ... 6’03 July to September ... ... 18 . .. 30 ... 7’03 October to December ... 16 . .. 24 ... 3’03 The correspondence is marked. We have now seen that, considering temperature and humidity together in their influence on the drying of mine roofs, we get correspondence both annually and seasonally, with accidents from falls of roof just as we did between rainfall and such accidents. The corre- spondence is, however, much more likely to be one of cause and effect. Indeed, it is what would be expected, for if we remember that in the collieries under con- sideration the roofs under which men work are nearly always either shale or impure coal, and that such materials contract when dried, we shall expect an increased absorption of water by the air-current to increase jje number of roof falls. If the roofs were of sandstone drying should have relatively little effect, and for Natal collieries, where that is generally the case, readily available statistics for the years 1905 to 1909 show 58 accidents from falls of roof in the months April to September and 54 in the other months of the year. To the writer it seems that the foregoing investigation fairly leads to the conclusion that though many things affect accident rates from falls of roof on the Middel- burg coalfield, the influence of the weather runs through those rates like the tide through disturbances of the sea. It is even possible to form some crude idea of the extent of that influence, for, if we take the last preceding table, and assume that the number of accidents in any period is made up of a constant plus a multiple of the relative drying figures, calculation indicates that approximately eight in the summer months and 23 in the winter months, making in all 31 out of the 105 accidents, or, roughly, 30 per cent., were due to weather conditions. Similarly, dealing with the previous table, a formula, number of accidents equals 6* * * §2 + 0’58 times relative drying figure, is deduced as giving a monthly approximation, and the results are plotted in diagram 2. That drying causes certain roof falls is well known, but whether or not changes in drying effect, due to atmospheric changes, have been recognised as having important influence on accidents is doubtful. The effect of seasonal deposit of moisture has, however, been remarked upon. For instance, in A. R. Sawyer’s Accidents in Mines, page 86, the following, passage occurs: in the same Way atmospheric vapour con- densed by coming in contact with strata colder than the air which conveys it may have a loosening effect on them. This is more likely to be the case in summer with shallow mines and unextended workings.” In South Africa atmospheric conditions can seldom cause deposit of moisture on the roof near the working Some Results of Low Temperature Carbonising Experiments. Wood. Brown lignite. Black lignite. Illinois coal. Kanawha gas. Smoke- less. Anthra- cite. Analyses (per cent.): H2O and volatile 75 70’5 65’6 42’0 29’0 20’0 7-9 Fixed carbon . 25 29’5 35’0 58’0 71’0 80’0 92’1 Temp, decomposition (degs. Cent.) 150 230 260 300 330 330 340? Temp, range of expts, (degs. Cent.) 150-188 230-320 260-340 300-330 300-360 330-400 230-475 Litres gas evolved from 500 grammes 43’5 10’8 14’0 11’5 14’0 9’3 4’6 Cubic feet gas per t >n 3,080 750 1,000 815 1,000 659 326 Gas analyses (per cent.): CO2 58’6 55’6 49’3 24’1 7’3 6’5 10’3 c6h0 0’5 0’6 0’9 2’8 0’5 2’6 1’3 <'.■11 0’6 1’0 0’8 2’2 1’4 2’3 0’4 co 35’0 24’7 19’5 10’6 6’0 4T 2’6 ch4 1’3 11’4 19’5 43’3 57’4 53’5 70’6 h2 1’0 0’3 0’0 3’4 14’6 9’7 0’0 N2 3’0 6’4 10’0 13’6 12’8 21’3 14’8 Per cent, by weight volatilised 51’9 26’8 31’0 23’7 12’2 10’0 0’7 Analysis of residue : Volatile 56’0 39’0 30’8 23’3 18’7 13’9 7’1 Fixed carbon 44’0 61’0 69’2 76’7 81’3 86T 92’9 face, as that can occur only in hot moist weather and when the face is near the mine entrance. Generally, such condensation as there may be will be on the intake airways. On the other hand, the air-current will rarely reach the working face so warm and charged with water vapour as to be saturated at the temperature of the roof and therefore unable to exert a drying influence. That being so, it is natural that occasionally a sheet of roof shale or a block of roof coal, which has been sounded and examined without detection of any open crack, should fall not long afterwards. Such things will occur most frequently when the drying effect is greatest, which is in cold dry weather, and more particularly on winter nights. If managers and men recognise the weather effect they will apply their own safeguards, but to the writer the practical lessons, particularly for winter guidance, seem to be—(a) sound and examine newly- exposed roof frequently, or prop it systematically; (b) prop the under side of every slip, which, if opened, may cause a fall; and (c), when practicable, course the air so that areas of bad roof shall be on the return side. If those things are done, a future investigator may fail to find any indication of weather influence on accidents from falls of roof in the Middelburg coalfield. SOUTH STAFFORDSHIRE AND WARWICKSHIRE INSTITUTE OF MINING ENGINEERS. A general meeting of the above institute was held at the University, Birmingham, June 29, the President (Dr. J. Cadman) in the chair.—The Secretary read the minutes of the last meeting, which were confirmed. The following gentlemen having been approved by the council were elected as members:— Prof. W. S. Boulton, D.Sc., the University, Birmingham; Mr. S. A. Matthews, Billingsley, Bridgenorth; Mr. S. R. Rao, Department of Industries, Bangalore, India.—A vote of congratulation, proposed by the President, was passed to Sir R. A. S. Redmayne, K.C.B., a past-president of this institute, upon the honour of knighthood recently conferred upon him. The President called upon Mr. Turquand to read his paper on “ An Instrument for the Detection of Firedamp in Mines.” At the suggestion of the President the publication of this paper was postponed, pending the addition of further tests. The President proposed a vote of thanks to Mr. Turquand, which was seconded by Mr. G. M. Cockin, and carried. Mr. Turquand briefly thanked the members. COAL AND ITS BY-PRODUCTS * By L. C. Jones.f Since commercial carbonising of coal gives the benzene series of products, and not paraffines as existing in petroleum and natural gas, efforts have been made to explain the origin of natural gas and petroleum from animal remains, and Moissan has suggested the carbide theory. We believe, however, that the slow distillation of coal at low temperatures is a more logical expla- nation.! Industrial experiments recently carried on in England upon low temperature distillation by the coalite process, and more recently on coking coals at even lower tem- peratures, and in vacuo §, indicate that petroleum pro- ducts identical with those found in nature are thus obtained. Low Temperature Carbonising Experiments. Recently,'in our laboratory at Syracuse, we have tried to determine the nature of the products at low tem- peratures, starting with the wood and going through the series to coke. Our experiments are still incomplete, but some of the results are tabulated below. The results are calculated to the pure, dry coal basis. In these experiments we used a cast-iron bomb, heated electri- cally or in a muffle furnace. This bomb, with absorbing apparatus and gas holder, was devised by the writer about eight years ago, and has been much used in the laboratories of the U.S. Government, and in our laboratories, for determining the nature and amount of products from various coking coals under varying conditions. The residue from the first carbonising test on the brown lignite was again heated from 450 to 500 degs. cent., producing 3,500 cu. ft. per ton of gas of analysis as shown under heading A. This second residue, heated again to 920 degs. cent., gave 7,000 cu. ft. of gas of analysis under heading B. A. Per cent. B. Per cent. CO2 27’2 7’0 c6h6 5’5 1’2 C2H4 3’2 IT o: 0’7 0’5 2 • ’ • CO 9’2 ... 22’5 ch4 38’4 ... 18’3 H ., 120 ... 41’8 n.; 3’8 ... 7’5 Evidently from these tests decomposition gradually brought about by heat tends to drive off at first the high oxygen content, and each coal type is converted into a type similar to, though not exactly like the succeeeding higher grade fuel. In the case of wood, the gas first evolved contains about 97 per cent, of CO2 and CO. Similarly the lignites and Illinois coals first give off a high percentage of oxy- genated compounds, and in this way necessarily the residues contain less oxygen and are of higher calorific value. Under high pressures doubtless a sharper frac- tional decomposition would result, and the residues even more closely resemble- the natural fuels. At any rate such slow, artificial carbonisation is strikingly sugges- tive of nature’s process. These residues show decided resemblances to the natural product. That they are not identical, both physically and chemically, may be due to the high pressure used in nature which we have made no effort to duplicate.U But if natural gas, chiefly methane and hydrogen, is a by-product of nature’s carbonising processes, we must explain the disappearance of the large amount of carbon dioxide and carbon monoxide which was originally pre- sent in such decomposition gases. May we not assume that the CO2 has been absorbed from the gas by the * Extracted from a paper read before the Franklin Institute. + Chief Chemist, Solvay Process Company, Syracuse. New York. X Mabery’s study of “Mahone Petroleum, its Recent Origin and the Origin of Petroleum in General,” has just come to my attention. The evidence given leaves little doubt that petroleum is one of nature’s by-products of coal. —J. I, and Eng. Chem., February 14, p. 101. § Pictet and Bouvier, Ber., 1913, p. 3342. T Results obtained by Ralston, United States Bureau of Mines, when plotting the composition of coals by the ingenious triangle method, indicate that the curve of decomposition bends strongly as semi-anthracite is approached, corresponding to the bend in our curve for calorific value of volatile matter.