Ironwood: IELTS Academic Reading practice test

Iron is the fourth most common element in the earth's crust, and almost none of it is metal. It occurs as oxide, chemically bound to oxygen, and the whole history of the iron industry is the history of methods for persuading the oxygen to leave. The reagent has almost always been carbon, which at a sufficient temperature takes oxygen away from iron and departs as gas. What has changed, repeatedly and consequentially, is how the heat and the carbon are delivered.

The earliest method was the bloomery, a shaft perhaps a metre high, filled in layers with ore and with charcoal and blown with hand or foot bellows. It was not hot enough to melt iron and was never meant to be. What emerged instead was a spongy mass called a bloom, iron mixed through with the stony waste known as slag, which had to be reheated and hammered repeatedly until most of the slag was driven out. The result was wrought iron, a soft and fibrous metal that can be forged into shape and welded, and each firing produced a few kilograms after a day of very hard work.

The blast furnace changed the scale by changing the temperature. A taller shaft, with air forced in by bellows driven from a waterwheel, reaches a point at which the iron does melt, and molten iron in contact with charcoal absorbs a great deal of carbon. The product runs out as a liquid and can be poured into a mould, which is an enormous practical advantage, but with several per cent of carbon in it the metal is hard, brittle and impossible to forge. Furnaces of this kind were working in China by the first century before our era; they appear in Europe rather more than a thousand years later.

Cast iron therefore created a second problem, which was how to turn it back into something that could be worked. The answer for centuries was to remelt it in a hearth with a blast of air playing on the surface, burning off the carbon by hand and by eye. The process that industrialised this was puddling, introduced in the 1780s, in which the metal is melted in a furnace so arranged that the flame never touches it and a workman stirs the bath with a long rod until the carbon has gone and the iron becomes pasty and can be lifted out. A generation earlier the industry had been freed from a different constraint when coke, made by baking coal, was successfully substituted for charcoal in the blast furnace. Before that every ton of iron had consumed a small woodland.

Steel sat between the two. It is iron with a controlled and small proportion of carbon, harder and more elastic than wrought iron and far tougher than cast, and before the middle of the nineteenth century it could only be made in tiny quantities. The best of it was produced by sealing wrought iron and charcoal in a clay crucible and holding it at a high temperature for hours, a method that yielded a few kilograms at a time at a price that confined steel to razors, springs, tools and swords. Bridges and rails were built of iron because steel was not available in tons.

In 1856 Henry Bessemer described a converter in which air is blown through molten pig iron from below. The carbon and the silicon in the metal burn, and the heat given off by that combustion is enough to keep the charge molten without any external fuel at all, so that a process which had taken a day took about twenty minutes and consumed nothing but air. The invention very nearly failed. It worked only with ores low in phosphorus, which in practice meant it worked in Sweden and in a few English districts and nowhere else in Europe; the difficulty was removed in 1878 when two cousins, a clerk and a chemist, showed that lining the converter with a basic material would take the phosphorus out into the slag.

The consequences are hard to overstate. The price of steel fell by something like ninety per cent within thirty years, and with it came rails that did not shatter in winter, ships built of plate, frames that let buildings rise past the height masonry allows, and the whole apparatus of machine tools and reinforced concrete. Two later inventions complete the story: the oxygen converter of the 1950s, which does Bessemer's job faster with pure oxygen instead of air, and the electric arc furnace, which makes steel by melting scrap and now accounts for a large share of world production.

Scrap, however, cannot supply the whole demand, since the steel in use is still growing and a beam put up today will not be available for fifty years. That leaves the primary route, and the primary route is a chemical reaction that requires carbon, which is why steelmaking accounts for a substantial share of global industrial emissions. The leading alternative is to reduce the ore with hydrogen instead, producing water rather than carbon dioxide. It works; the difficulty is that making the hydrogen would require quantities of clean electricity comparable to the entire present output of several countries, and that a blast furnace lasts decades, so the plants being relined this year will still be running in the 2050s.

Anyone who follows environmental reporting will have met a figure of the following kind: that the world is losing dozens of species every day, or one every few minutes, or a hundred times as many as it should be. The numbers are quoted without qualification and are seldom sourced. They are not invented, and they are not measurements either. They are the output of a chain of estimates, each defensible on its own and each carrying an uncertainty that the final figure does not display.

Begin with what is actually known. Somewhat over two million species have been formally described and named. How many exist in total is genuinely unsettled: published estimates have ranged from about two million to over a hundred million, and a widely used figure of roughly nine million rests on an extrapolation from the pattern by which higher groupings have been filled in over two centuries. As for extinctions, the number of species documented as having disappeared since 1500 is in the region of nine hundred. Everybody in the field agrees that this figure is far too low, and the disagreement is about by how much.

The large numbers come from indirect methods. The most important rests on a long-observed regularity: the number of species found in an area rises with the area in a predictable way, so that shrinking a habitat to a tenth of its extent should, on this relationship, cost roughly half of its species. Applying the curve backwards to measured habitat loss yields a figure for species committed to extinction, which is then scaled up to cover the groups nobody has surveyed. A second method compares the current documented rate with the rate inferred from the fossil record and reports the ratio, which is where the multiples of a hundred or a thousand come from.

Both methods have been attacked, and one attack in particular deserves attention. It has been argued that using the species-area relationship in reverse is mathematically invalid, because the area required to find the first individual of a species is not the same as the area required to remove the last one; by this account the standard method overestimates substantially. There are further problems of a more familiar kind. Our knowledge is concentrated in birds and mammals, which are perhaps two per cent of animal species and the great majority of confirmed extinctions. And declaring an extinction is a judgement that can be reversed: species pronounced gone have been rediscovered decades later, which makes assessors reasonably cautious about declaring the next one.

That caution cuts the other way as well, and this is where the critics tend to stop too soon. If most species are undescribed, their disappearance cannot appear in any register by definition, and the groups least studied are exactly the invertebrates and plants of tropical forests, where habitat loss has been heaviest. Land snails, to take a well-documented case, have almost certainly lost several hundred species, of which a handful appear in the official lists. More importantly, extinction is the last event in a long sequence, and the sequence is visible well before its end. Populations contract, ranges shrink, and a species may be functionally gone from ninety per cent of its former territory while remaining, in the counting, entirely present.

This is why the argument about the headline number seems to me to be doing more harm than good. A figure that is precise, memorable and unsupported is a liability, because it will eventually be examined. When it is, and when someone demonstrates that the true documented rate is a tenth of what was claimed, the demonstration is reported as though it had settled the underlying question, which it has not. The problem was never that a particular number was right. It is that the direction of travel is not in dispute and the magnitude cannot be pinned down, and those two facts are difficult to hold in the same sentence without sounding either alarmist or complacent.

There are better instruments than a global extinction count, and they are the ones specialists actually use. An index built from the movement of assessed species between threat categories measures whether things are getting worse and by how much, without requiring anybody to certify a death. Measures of range contraction do the same for geography. Indices of population abundance track the sequence that precedes extinction rather than its final moment, though the best known of them has attracted serious statistical criticism for the weight a small number of collapsing populations can exert on the average. None of these produces a sentence as quotable as one species every twenty minutes, which is precisely their disadvantage in public.

My own position is that the honest statement is also the useful one, and it runs as follows. We do not know the rate of extinction and are unlikely to know it, because we do not know what is there to be lost. We do know that habitat is being removed faster than it is being replaced, that assessed populations are declining on average, and that the pattern is consistent across every group that has been examined properly. The uncertainty is about magnitude and not about sign. Waiting for a reliable number before acting is not caution, because the number would arrive, if it ever did, from a census conducted after the fact.