economically than a good quality of rail can be produced in iron from the cheaper ironstone of Cleveland or elsewhere, steel, to the entire exclusion of iron, must be henceforward looked upon as the proper material for railroads. Iron, in the matter of wear, exhibited very great irregularity, some rails showing signs of distress within a year or two of being laid down, while others afforded very satisfactory results. This uncertainty of quality has perhaps been the means of procuring for iron a worse reputation than its merits as a whole deserved.

In illustration of this assertion I would instance the experience on the main line of the North-Eastern Railway on certain sections of its system, which may be taken as fair samples of the others. On that extending between Newcastle and Berwick, 66.8 miles of double way, the iron rails laid down in 1847 weighed 65 lbs. per yard. Renewals commenced in 1855 and terminated in 1867. In these the weight was increased to 82 lbs. per yard.

The maximum duration of the 65 lbs. rails was 21 years and the minimum 8 years, the average being 12.8 years.

A second relaying with the 82 lbs. rails was commenced in 1864, and the average age of those taken out has been 14:18 years. The iron rails still in use on the Newcastle and Berwick section is 37.30 miles of single road, having a present (June, 1883) average age of 14.87 years.

The Newcastle and Darlington section has a length of 35 15 miles of double way. Rails of 65 lbs. per yard, laid down in 1844, began to be renewed in 1853 and others of 82 and 83 lbs. were laid in their place. The longest life of those of 65 lbs. was 19 years and the shortest 9 years the average of the whole being 11:47 years. The average life of the heavier rails (82 and 83 lbs.) was only 9.38 years, but the traffic became much heavier after they were brought into use. There only remains 3:32 miles of single way of iron rails on this section, the average present life of which is 8:40 years.

The Newcastle and Carlisle section comprises 59.4 miles of double way. It was laid with Losh's patent "fish bellied rails," weighing only 42 lbs. per yard, partly rolled at Walker-on-Tyne but chiefly in South Wales. The maximum duration of these was 26 years and the minimum 13 years, the average being 17.25 years.

The average life of 83 lbs. rails, commenced to be laid down on this section in 1849 was 14-21 years. There still remains (June, 1883) 69.2 miles (single way) of iron rails on this section, the average age of which is 10:05 years.

The statements just submitted do not afford any proper criterion of the resisting powers of iron rails; for this can only be determined by the comparative weights of the engines, the amount of traffic and the speed of the trains which have passed over them. According to Mr. R. Price Williams the average life of an iron rail, on the most heavily worked portions of the railways in the United Kingdom, may roundly be taken at about 17 millions of tons. Quoting the experience of Mr. Webb of the London and North-Western Railway, Mr. Williams proceeds to estimate that the average life of a Bessemer steel rail will be 161 millions of tons, or approximately about nine times that of an iron rail. This was an opinion expressed about the year 1878, but it is to be apprehended that actual experience, in which even now (1883) we are greatly deficient, will not be found to corroborate this opinion of the relative durability of the two kinds of rails.

On the North-Eastern system comprising 1,508 miles of main line of railway equal to 2,499 miles of single line, steel rails were only introduced, at first very sparingly, about the year 1862 and it was not until 1874 that they were largely used. Since the year 1877 no iron rails have been purchased by this company. At the present moment 1,303 single miles of their way are laid with steel and the quantity renewed as not fit for further use, is too insignificant to afford any data upon which to form any opinion of their durability. The only accurate data in my possession is that afforded by 1,216 yards in the Shildon tunnel, on the Stockton and Darlington Section, having a gradient of 1 in 250 against the load. In June and July, 1865, this length was laid with double-headed steel rails weighing 77 lbs. per yard, of which the wear was ascertained from time to time.

By March, 1869, the average loss of weight on twenty-four rails was found to be 8 lbs. per yard, equal to 10.39 per cent., an amount of abrasion which necessitated their being turned.

In June, 1872, an average of twenty-three rails showed a further wear of 9 lbs. per yard equal to 11.69 per cent., making 22.08 per cent.

during the seven years they had been in use, equal therefore to 3.15 At this period the whole were taken up as being

per cent. per annum.

no longer serviceable.

The total net traffic which passed through the tunnel was 31,149,473 tons, or including the weight of the trucks, etc., say 46 million tons. Roughly we may consider the return load would be that of the empty trucks, etc., say 15 million tons. Taking the entire weight which passed over the two lines of way we have 614 millions, which would give 30.62 million tons as the life of a rail in this particular tunnel. I am not prepared to say that the wear of a rail in such a situation is the same as that in the open air; but I do know that one-ninth of the period during which they were in use, viz., year, or less than ten months, did not express the life of an iron rail in this locality, which it ought to do according to the formula laid down by Mr. Price Williams.

So far as the experience of the officials of the North-Eastern Railway enables them to judge at present, they will be well satisfied if the steel rails now in use have double the life of those of iron of the same weight per yard (82 lbs.) which preceded them, i.e., instead of remaining serviceable for ten to twelve-and-a-half years they are found in use at the end of twenty-two years, calculated on the present traffic.

Whatever the power of steel rails to resist abrasion may be, all experience points to their life being much more uniform than was the case where iron was the material employed. An iron rail however became unserviceable long before it had lost much of its weight. Instead of actual wear, the head of the rail became crushed, its comparative toughness apparently resisting abrasion. Thus while the average loss of weight in iron rails would not exceed 7 per cent. at the period of removal, the loss in 82 lbs. steel rails appears to be 1 lb. per yard per annum-say 14 per cent. for each year of actual service.

Before economy in the manufacture of steel rails had reached its present position, the North-Eastern Railway Company tried iron, in the form of a solid bloom, obtained from the pig of their own district in the revolving furnace. The puddled iron was treated in machinery constructed by Sir W. G. Armstrong & Co., and the rails were casehardened by Dodds' process. The change in the cost of producing steel rails however put an end to further trials in this direction.

SECTION XIV.

ON MORE RECENT METHODS OF SEPARATING THE SUBSTANCES TAKEN UP BY IRON DURING ITS PASSAGE THROUGH THE BLAST FURNACE.

In a former portion of this work the advantages were set forth of the circuitous mode of treating the ores of iron by the aid of the blast furnace, even when the metal in its malleable state was the ultimate object in view. When iron is required in this last mentioned form, the manufacturer has hitherto been unable to obtain it in large masses, by the ordinary method of simple fusion, as pursued by smelters of other metals in common use, such as gold, silver, copper, lead, or zinc. The melting point of wrought iron is so high, that it is only within the last quarter of a century that we have been able to bring any quantity, beyond a few pounds, to the fluid state at one time, by means of heat. Under these circumstances the iron maker had to avail himself of that other valuable property possessed by this metal, namely welding; by the aid of which his granules were built up into the shape of the "puddler's ball" by the manual exertion of the workThis required extensive shops containing squeezers, hammers, puddle-rolls, etc., by means of which, and of a considerable expenditure of metal, fuel, and labour, the puddled iron was welded together. The product thus obtained is liable to unsoundness, from the difficulty of separating the adhering cinder, which, as is well known, gets sealed up in the interior of the mass.

man.

In 1856 Henry Bessemer announced that he was able to obtain fluid malleable iron by blowing air through the molten crude metal employed in its production, the heat being obtained by the combustion

of those substances which had united themselves with the iron in the blast furnace. In other words the very heat, as it were, which had effected the reduction of the metalloids, in the smelting process, was rendered available in the Bessemer converter. The surprise such a statement was calculated to excite was a good deal subdued by the small amount of faith with which it was received.

No doubt it did seem at first sight incredible that such an effect as that represented could be obtained in any form of apparatus; seeing that, in the old fashion of making malleable iron by puddling, the metalloids contained in the pig were equally oxidised, with the same evolution of heat as that generated in the converter. The result with the latter was the more surprising, inasmuch as in that case there is an enormous excess of heat, which escapes in the form of intensely brilliant flame from the mouth of the vessel; while in puddling the metal a quantity of coal has to be burnt, most wastefully it is true, equal to half that required in the smelting furnace.

In forming a judgment on the subject the difference of the conditions was entirely overlooked. In the puddling furnace the internal heat was produced by the comparatively slow combustion of the carbon and slow oxidation of the metalloids; while the heat was applied externally to the raw material in an apparatus from which there was a great loss by radiation, convection, etc. In the converter, on the other hand, an immense volume of air was rapidly poured through the metal, generating the heat in the very heart of that mass the temperature of which it was desired to raise. The operation was also performed under circumstances where the loss by radiation, having regard to the quantity under treatment, was infinitely less than in an ordinary furnace.

For some years after the first brilliant exhibitions of a Bessemer converter at work, those interested in its success encountered great disappointment. The fact that atmospheric air, when forced into molten cast iron, would burn off carbon, silicon, and sulphur, and would partially remove the phosphorus, was well known to every one who had any experience in the violent action set up towards the close of the old process of refining. No one suspected however, until Bessemer proved it, that by forcing the air upwards through the mass of liquid iron in sufficient quantity, and with sufficient velocity, so