the Boghead coal of Linlithgowshire, Scotland, an analysis of which by Dr. Penny is as follows: Proximate-moisture 0.84; vol. 67.95; fixed C, 9.54, ash, 21.4; Ultimate-C,63.94; H, 8.86; O, 4.70; N, 0.96; which is remarkable for the high percentage of H.

*The analyses of Iowa, Missouri, Nebraska, and Wyoming coals are selected from a paper on The Heating Value of Western Coals, by Wm. Forsyth, Mech. Engr. of the C., B. & Q. R. R., Eng'g News, Jan. 17, 1895.

Includes sulphur, which is very high. Coke from Cedar City analyzed: Water and volatile matter, 1.42; fixed carbon, 76.70; ash, 16.61; sulphur, 5.27.

Colorado Coals.-The Colorado coals are of extremely variable composition, ranging all the way from lignite to anthracite. G. C. Hewitt (Trans. A. I. M. E., xvii. 377) says: The coal seams, where unchanged by heat and flexure, carry a lignite containing from 5% to 20% of water. In the south-eastern corner of the field the same have been metamorphosed so that in four miles the same seams are an anthracite, coking, and dry coal. In the basin of Coal Creek the coals are extremely fat, and produce a hard, bright, sonorous coke. North of coal basin half a mile of development shows a gradual change from a good coking coal with patches of dry coal to a dry coal that will barely agglutinate in a beehive oven. In another half mile the same seam is dry. In this transition area, a small cross-fault makes the coal fat for twenty or more feet on either side. The dry seams also present wide chemical and physical changes in short distances. A soft and loosely bedded coal has in a hundred feet become compact and hard without the intervention of a fault. A couple of hundred feet has reduced the water of combination from 12% to 5%.

Western Arkansas and Indian Territory. (H. M. Chance, Traus. A. I. M. E. 1890.)-The Choctaw coal-field is a direct westward exten

sion of the Arkansas coal-field, but its coals are not like Arkansas coals, except in the country immediately adjoining the Arkansas line.

The western Arkansas coals are dry semi-bituminous or semi-anthracitic coals, mostly non-coking, or with quite feeble coking properties, ranging from 14% to 16% in volatile matter, the highest percentage yet found, according to Mr. Winslow's Arkansas report, being 17.655.

In the Mitchell basin, about 10 miles west from the Arkansas line, coal recently opened shows 19% volatile matter; the Mayberry coal, about 8 miles farther west, contains 23% volatile matter; and the Bryan Mine coal, about the same distance west, shows 26% volatile matter. About 30 miles farther west, the coal shows from 38% to 41%% volatile matter, which is also about the percentage in coals of the McAlester and Lehigh districts.

Western Lignites. (R. W. Raymond, Trans. A. I. M. E., vol. ii. 1873.)

deducting the heat required to vaporize the moisture and combined water, that is, 537 calories for each unit of water. 1 calorie = 1.8 British thermal anits.

Analyses of Foreign Coals. (Selected from D. L. Barnes's paper on American Locomotive Practice, A. S. C. E., 1893.)

An analysis of Pictou, N. S., coal, in Trans. A. I. M. E., xiv. 560, is: Vol.. 29.63; carbon, 56.98; ash, 13.39; and one of Sydney, Cape Breton, coal is: vol., 34.07; carbon, 61.43; ash, 4.50.

Nixon's Navigation Welsh Coal is remarkably pure, and con. tains not more than 3 to 4 per cent of ashes, giving 88 per cent of hard and lustrous coke. The quantity of fixed carbon it contains would classify it among the dry coals, but on account of its coke and its intensity of combustion it belongs to the class of fat, or long-flaming coals.

Chemical analysis gave the following results: Carbon, 90.27; hydrogen, 4.39; sulphur, .69; nitrogen, .49; oxygen (difference), 4.16.

The analysis showed the following composition of the volatile parts: Carbon, 22.53; hydrogen, 34.96; O+N+S, 42.51.

The heat of combustion was found to be, as a result of several experiments, 8864 calories for the unit of weight. Calculated according to its composition, the heat of combustion would be 8805 calories = 15,849 British thermal units per pound.

This coal is generally used in trial-trips of steam-vessels in Great Britain. Sampling Coal for Analysis.-J. P. Kimball, Trans. A. I. M. E., xii. 317, says: The unsuitable sampling of a coal-seam, or the improper preparation of the sample in the laboratory, often gives rise to errors in determinations of the ash so wide in range as to vitiate the analysis for all practical purposes; every other single determination, excepting moisture, showing its relative part of the error. The determination of sulphur and ash are especially liable to error, as they are intimately associated in the slates.

Wm. Forsyth, in his paper on The Heating Value of Western Coals (Eng'g News, Jan. 17, 1895), says: This trouble in getting a fairly average sample of anthracite coal has compelled the Reading R. R. Co., in getting their samples, to take as much as 300 lbs. for one sample, drawn direct from the chutes, as it stands ready for shipment.

The directions for collecting samples of coal for analysis at the C., B. & Q. laboratory are as follows:

Two samples should be taken, one marked "average," the other "select." Each sample should contain about 10 lbs., made up of lumps about the size of an orange taken from different parts of the dump or car, and so selected that they shall represent as nearly as possible, first, the average lot; second, the best coal.

'An example of the difference between an sample, taken from Mr. Forsyth's paper, is the Moisture. Vol. Mat.

average and a "select" following of an Illinois coal: Fixed Carbon.

35.41
48.23

Ash.

35.54

15.17

The theoretical evaporative power of the former was 9.13 lbs. of water from and at 212° per lb. of coal, and that of the latter 11.44 lbs.

Relative Value of Fine Sizes of Anthracite.-For burning on a grate coal-dust is commercially valueless, the finest commercial anthracites being sold at the following rates per ton at the mines, according to a recent address by Mr. Eckley B. Coxe (1893):

Barley...

........

Range of Size.
1% to % inch

% to 9/16

9/16 to 3%
3% to 3/16
3/16 to 2/32

But when coal is reduced to an impalpable dust, a method of burning it becomes possible to which even the finest of these sizes is wholly unadapted; the coal may be blown in as dust, mixed with its proper proportion of air, and no grate at all is then required.

Pressed Fuel. (E. F. Loiseau, Trans. A. I. M. E., viii. 314.)-Pressed fuel has been made from anthracite dust by mixing the dust with ten per cent of its bulk of dry pitch, which is prepared by separating from tar at a temperature of 572° F. the volatile matter it contains. The mixture is kept heated by steam to 212°, at which temperature the pitch acquires its cementing properties, and is passed between two rollers, on the periphery of which are milled out a series of semi-oval cavities. The lumps of the mixture, about the size of an egg, drop out under the rollers on an endless belt which carries them to a screen in eight minutes, which time is sufficient to cool the lumps, and they are then ready for delivery.

The enterprise of making the pressed fuel above described was not commercially successful, on account of the low price of other coal. In France, however, briquettes' are regularly made of coal-dust (bituminous and semi-bituminous).

RELATIVE VALUE OF STEAM COALS.

The heating value of a coal may be determined, with more or less approx. imation to accuracy, by three different methods.

1st, by chemical analysis; 2d. by combustion in a coal calorimeter; 3d, by actual trial in a steam-boiler. The first two methods give what may be called the theoretical heating value, the third gives the practical value.

The accuracy of the first two methods depends on the precision of the method of analysis or calorimetry adopted, and upon the care and skill of the operator. The results of the third method are subject to numerous sources of variation and error, and may be taken as approximately true only for the particular conditions under which the test is made. Analysis and calorimetry give with considerable accuracy the heating value which may be obtained under the conditions of perfect combustion and complete absorption of the heat produced. A boiler test gives the actual result under conditions of more or less imperfect combustion, and of numerous and variable wastes. It may give the highest practical heating value, if the condi tions of grate-bars, draft, extent of heating surface, method of firing, etc., are the best possible for the particular coal tested. and it may give results far beneath the highest if these conditions are adverse or uusuitable to the coal.

The results of boiler tests being so extremely variable, their use for the purpose of determining the relative steaming values of different coals has frequently led to false conclusions. A notable instance is found in the record of Prof. Johnson's tests, made in 1844, the only extensive series of tests of American coals ever made. He reported the steaming value of the Lehigh Coal & Navigation Co.'s coal to be far the lowest of all the anthracites, a result which is easily explained by an examination of the conditions under which he made the test, which were entirely unsuited to that coal. He also reported a result for Pittsburgh coal which is far beneath that now obtainable in every-day practice, his low result being chiefly due to the use of an improper furnace.

In a paper entitled Proposed Apparatus for Determining the Heating Power of Different Coals (Trans. A. I. M. E., xiv. 727) the author described and illustrated an apparatus designed to test fuel on a large scale, avoiding the errors of a steam-boiler test. It consists of a fire-brick furnace enclosed in a water casing, and two cylindrical shells containing a great number of tubes, which are surrounded by cooling water and through which the gases of combustion pass while being cooled. No steam is generated in the apparatus, but water is passed through it and allowed to escape at a temperature below 200° F. The product of the weight of the water passed through the apparatus by its increase in temperature is the measure of the heating value of the fuel.

There has been much difference of opinion concerning the value of chemical analysis as a means of approximating the heating power of coal. It was found by Scheurer-Kestner and Meunier-Dollfus, in their extensive series of tests, made in Europe in 1868, that the heating power as determined by calorimetric tests was greater than that given to chemical analysis according to Dulong's law.

Recent tests made in Paris by M. Mahler, however, show a much closer agreement of analysis and calorimetric tests. A brief description of these tests, translated from the French, may be found in an article by the author in The Mineral Industry, vol. i. page 97.

Dulong's law may be expressed by the formula,

Heating Power in British Thermal Units

14,500C + 62,500 (H
(H - ),*

in which C, H, and O are respectively the percentage of carbon, hydrogen, and oxygen, each divided by 100. A study of M. Mahler's calorimetric tests shows that the maximum difference between the results of these tests and the calculated heating power by Dulong's law in any single case is only a little over 3%, and the results of 31 tests show that Dulong's formula gives an average of only 47 thermal units less than the calorimetric tests, the average total heating value being over 14,000 thermal units, a difference of less than 4/10 of 1%.

* Mahler gives Dulong's formula with Berthelot's figure for the heating value of carbon, in British thermal units,

14,650C + 62,025 (H
(H - (0 +N) - 1).

Mahler's calorimetric apparatus consists of a strong steel vessel or "bomb" immersed in water, proper precaution being taken to prevent radiation. One gram of the coal to be tested is placed in a platinum boat within this bomb, oxygen gas is introduced under a pressure of 20 to 25 atmospheres, and the coal ignited explosively by an electric spark. Combustion is complete and instantaneous, the heat is radiated into the surrounding water, weighing 2200 grams, and its quantity is determined by the rise in temperature of this water, due corrections being made for the heat capacity of the apparatus itself. The accuracy of the apparatus is remarkable, duplicate tests giving results varying only about 2 parts in 1000.

The close agreement of the results of calorimetric tests when properly conducted, and of the heating power calculated from chemical analysis, indicates that either the chemical or the calorimetric method may be accepted as correct enough for all practical purposes for determining the total heating power of coal. The results obtained by either method may be taken as a standard by which the results of a boiler test are to be compared, and the difference between the total heating power, and the result of the boiler test is a measure of the inefficiency of the boiler under the conditions of any particular test.

In practice with good anthracite coal, in a steam-boiler properly proportioned, and with all conditions favorable, it is possible to obtain in the steam 80% of the total heat of combustion of the coal. This result was nearly obtained in the tests at the Centennial Exhibition in 1876, in five different boilers. An efficiency of 70% to 75% may easily be obtained in regular practice. With bituminous coals it is difficult to obtain as close an approach to the theoretical maximum of economy, for the reason that some of the vola tile combustible portion of the coal escapes unburned, the difficulty increas ing rapidly as the content of volatile matter increases beyond 20%. With most coals of the Western States it is with difficulty that as much as 60% or 65% of the theoretical efficiency can be obtained without the use of gas-producers.

The chemical analysis heretofore referred to is the ultimate analysis, or the percentage of carbon, hydrogen, and oxygen of the dry coal. It is found, however, from a study of Mahler's tests that the proximate analysis, which gives fixed carbon, volatile matter, moisture, and ash, may be relied on as giving a measure of the heating value with a limit of error of only about 3%. After deducting the moisture and ash, and calculating the fixed carbon as a percentage of the coal dry and free from ash, the author has constructed the following table:

APPROXIMATE HEATING VALUE OF COALS.

Below 50% the law of decrease of heating-power shown in the table apparently does not hold, as some cannel coals and lignites show much higher heating power than would be predicted from their chemical constitution. The use of this table may be shown as follows:

Given a coal containing moisture 2%, ash 8%, fixed carbon 61%, and volatile matter 29%, what is its probable heating value? Deducting moisture and ash we find the fixed carbon is 61/90 or 68% of the total of fixed carbon and volatile matter. One pound of the coal dry and free from ash would, by the table, have a heating value of 15,480 thermal units, but as the ash and moist. ure, having no heating value, are 10% of the total weight of the coal, the coal would have 90% of the table value, or 13,932 thermal units. This divided by 966, the latent heat of steam at 212° gives an equivalent evaporation per lb. of coal of 14.42 lbs.