considerable spaces which often separate the prolongation of the same bed.

Miners are sometimes very much embarrassed, when they have pursued the trace of a fault during a long space; for, not finding again any indication of resumption, they do not know whether the interruption should be attributed to the presence of a fault or to the definitive cessation of the coal. No rule can be laid down in this case; the observation of the structure of the whole can only furnish the data which we have already indicated above.

Faults are very common accidents. They are fractures which affect the entire character of the formation, and cause greater or less disturbances. These faults have a determinate direction, and frequently a basin is affected by a system of faults parallel with each other. At other times, there are several systems which follow different directions, but are each composed of faults, connected with each other, by a parallelism of direction.

The intensity of the faults is very variable; sometimes they scarcely interrupt the formation, and appear as fissures which have changed the level of the two ruptured parts, but not enough to constitute the total interruption of the coal, which is always easy to follow, when the offset is not greater than the thickness of the coal. The bed of Lucy, in the basin of Saône-et Loire, frequently presents such faults.

The section of the coal bed of Monceau furnishes an example of interruptions of continuity caused by movements subsequent to the formation of the coal deposit, and of a confusion [brouillage] which totally interrupts the coal.

The brouillages are nothing but the intervals comprised between the planes of fracture; in these intervals all the beds are broken and reduced to angular blocks mingled together.

When these faults form part of those which have determined the outline of the surface of the ground, the offsets or upthrows are, in some measure, proportionate to the inequalities which it presents. Thus, there are offsets and upthrows of several hundred feet in the basins of England, of Wales, and of Rive-de-Gier, whose surface is highly disturbed.

In the basins of the Saône-et-Loire, they are rarely more than from thirty to ninety feet, and are, notwithstanding, in agreement with the undulations of the surface.

Endeavours have been made to establish, from the most detailed and available information, agreements between the arrangement of the beds of coal and the superficial accidents of the soil. Thus, in a great number of basins, the direction of their beds coincide with that of the great axis of the coal formation, and this great axis is itself directed in the same course as the existing valleys, in such a manner that the direction of the beds. is confounded with that of the dividing lines of the waters and the principal valleys. In some other basins, the planes of stratification of the formation have not only the

same direction, but also the same inclination as that of the surface declivities.

The basin of Bert offers a striking example of this agreement. The inclinations of its coal beds change as often as five times whilst conforming to the inclinations of the surface, and the range of the valleys corresponds with the direction of the beds. The greater part of the basins which affect the boat form, present, at their surfaces, an analogous disposition; that is to say, the waters follow the direction of the strata, and the lateral margins of the basins are generally more elevated than the axes.*

English Coal-Fields.-Faults and interruptions prevail, more or less, as might be expected, in most coal-fields, but they possess different characters in different regions. The Newcastle coal-field is remarkable for its number of faults; from the dimensions of a few inches to a hundred fathoms. But in the southern coal basins, particularly those of the Forest of Dean and South Wales, there are frequently found remarkable irregularities, called "horses." Where these horses occur, the coal disappears all at once; but yet without any fault at all. They have to be cut through, and, after a time, the coal reappears.† These horses appear to be ascribable to interruptions in the original deposition of the vegetable matter of the coal

seams.

Chemical Geology, as applied to Coal.-At the tenth annual meeting of the British Association for the Advancement of Science, as reported in the Athenæum, Professor Johnston brought forward the result of his investigations on the most important of mineral productions, coal.

Although some geologists may entertain a different opinion, he assumes for granted the vegetable origin of coal. Although it may be classified in various ways, for economic or geologic convenience, as into caking or not caking, bituminous or non-bituminous, the true basis of the classification must depend on the chemical composition. Carbon, oxygen, and hydrogen, are the component parts of living vegetables, and the same elements compose coal, but in different proportions.

In the decomposition of vegetable matter, there are two agents always at work-viz. atmospheric air and water, which resolve it into carbon, oxygen, and hydrogen; forming, with one another, those combinations-carburetted hydrogen, carbonic acid, and water. Vegetable matter, consequently, in different states, showed different proportions of these elements.

The quantity of carbon in all the different varieties of coal, in Mr. Johnston's table, was taken as a constant quantity; and from lignite, downwards, we see a progressive loss of hydrogen and oxygen; until, in anthracite, the carbon is the chief component.

This is borne out by experience. In the change from lignite to

* Géologie appliquée, par M. Amédée Burat, 1846. † Professor Ansted's Geological Lectures, 1847-8.

fossil wood we find that carbonic acid only is parted with; and this continues, without variation, in all the kinds, down to cannel coal.

In mines of lignite and cannel coal, we find only carbonic acid, or choke damp; while in mines of coal lower in the scale, we find, in addition, carburetted hydrogen or fire damp. This also appears in the table referred to; the hydrogen diminishing in each variety as we approach anthracite.

In some mines we find a perfect confirmation of this theory. In certain Yorkshire mines, coal of different kinds, cannel coal being at the top, evidently prove that those below, having been longer subjected to chemical action, had parted with more of their hydrogen. The same occurs in mines in Lancashire.

In conclusion, Professor Johnston asserted, that bituminous matter must be of vegetable origin-in fact, chemistry proved it. Distillation of vegetable matter in a gas work, or in the laboratory of a volcano, was the same process.

In further support of this conclusion, we cite the following high authority:

Table of Analysis of Coal and certain allied Combustibles, by Berthier.

These different varieties of brown coal, peat, bituminous coal, anthracite, and graphite, correspond so exactly, that this alone would show the vegetable origin of them all; from the peat up to the graphite, if no other proofs were at hand. [See Table next page.]

In the belief that every species of information which makes the adaptation of the various mineral combustibles to the manufacture of iron better understood, must be useful and in strict conformity with the plan of the present work, we have arranged the following practical details. Great changes have taken place, within a few years, in the management of fuel, and in the degree of estimation in which each species is held by operative and scientific men. It is proper to know the conclusions to which those persons have arrived. We cannot here give all those results in detail; and, moreover, this is not a treatise on iron making. But we have sought to concentrate certain material facts on the nature and capabilities of the principal varieties. It will then be easy to compare them with others of corresponding character. We have therefore given, in the following page, a comparative table and characteristic analysis of the principal descriptions of coal employed in the iron works of Europe and the United States.

Varieties of Coal, with reference to their Adaptation to the making of Iron.

A, Coals which cannot be employed in iron works, in the crude state.

B, Coals which cement less in the fire, and which it is practicable to use raw in furnaces worked with heated air,

C, Chiefly for illuminating gas.

CLASSIFICATION OF MINERAL COALS.

In the foregoing table of analysis of coals and anthracites we have so arranged them as to exhibit their varieties or gradations, and their distinguishing properties, in different countries. Hence, the European coals can readily be compared with those of America, and the adaptations of either may be assigned with some degree of confidence. We proceed to note these characteristic differences and agreements more in detail.

I. Fat Bituminous, blazing, coking.-In the first class, series A, of the table, by way of illustration, the English coals of the north, and some of the coals of Silesia, of Hesse, of France, and of Ame

rica, in the Ohio Valley, are chiefly fat and very adhesive or caking; swelling much in the fire. The hot air blast is successfully applied with these in the high furnaces. But, as their tendency to cement together in a solid mass, when in the fire, is such as to prevent a free draft or passage of the air through the furnace, it has been found indispensable to submit the coals to a preliminary process, and to reduce them to coke. Thus, the difficulty is wholly removed; and a light, cellular, and purely carbonaceous substance, easily ignited, is substituted for the unmanageable coal in its crude state. The average quantity of carbon which the English coals possess, is stated to be sixty-five per cent.

Series B, more southward, in Staffordshire and Derbyshire: these coals, although containing as much, and even more bitumen, do not melt together like those of Northumberland. They scarcely change their form even in the state of coke. The varieties, having this property, admit of their being used in the raw state, but require the introduction of hot air into the furnaces. Some of the American coals west of the Alleghany mountain have also these characters.

In regard to the manufacture of illuminating gas, the type of perfection, in the series C, is the Scotch cannel; then comes after it the Lancashire cannel, and, in the third order, the Yorkshire and Derbyshire cannel. With this class we would place the cannel coals of Kentucky, Indiana, Illinois and Missouri. This series can be assimilated, in many respects, to the coal of the basin of Mons in Belgium. The splint coal of Scotland is only a coarse variety of cannel, as are the greatest part of all the Scotch coals.

The Newcastle coals have a resemblance to those of Anzin, of Saint-Etienne, and of Rivè-de-Gier, the analysis of which we have placed in the series D.

II. Series E.-In the second class, the Scotch coals, although containing as much bitumen as those of the north of Scotland, are of the kind denominated dry coals. They cement together, but without change of form, and are not so adhesive as the fat English coals. These were heretofore coked before being put into the furnaces; but recent improvements have shown that, with the application of heated air, they can be employed without being previously carbonized. Their average proportion of carbon is about sixty per cent., and of bitumen 36 per cent. Some of the Alleghany coals will probably be found to assimilate with these. Approximating to the same class, to a certain point, will be found the coals of Auvergne and of a part of the south of France.

III. Series F.-We have assigned an intermediate space for a series of coals in the American coal basins which differ little from E, except that they contain somewhat less of bitumen and more of carbon, viz. about 66 per cent. of carbon and 27 per cent. of bitumen and volatile matter, and are less adhesive and caking. The Heraclea coal, in Anatolia, appears to belong to this series, and those of the Cantal and Puy de Dôme in France. These are convertible into coke.