io6 



NATURE 



{June i, 1882 



phenomena of the conversion of iron into steel in the cementation 

 process all point to the conclusion that the carbon is simply 

 absorbed, as the varying rate of impregnation with variations of 

 temperature, the gradual change from the outside to the inside, 

 and the large deposition of free carbon from such steel, if dis- 

 solved in hydrochloric acid, or chloride of copper, or cold dilute 

 nitric acid. 



As to No. 2, the author adopts the the 517 of Jullien, that the 

 hardening of steel is due to the crystallisation of the so-called com- 

 bined carbon (really absorbed) in a form resembling the diamond. 

 He observes that cemented steel only becomes hard when 

 heated and quenched, and that the fracture then shows innumer- 

 able small crystals, which, under the microscope, present physical 

 features very much like small diamonds. These crystals do not 

 appear in wrought iron, increase in number as the proportion of 

 carbon increases, and as the hardening increases, and are more 

 numerous at the outside of the piece, where the hardness is also 

 less. They are therefore crystallised carbon, in other words, 

 diamond. Estimations of carbon in the different hyers of a 

 piece of hardened steel have always shown that the actual pro- 

 portions, as formed by combustion, are the same throughout, 

 but that, as examined by the colour test, they increase gradually 

 from the outside to the inside. This shows that some change 

 has taken place in the carbon. The author's theory is that at a 

 red heat the molecules of iron are expanded and partially sepa- 

 rated ; that in this state the absorbed carbon is partially disso- 

 ciated from the iron, and upon the steel being suddenly 

 quenched, the carbon is not re-absorbed, but takes up a small 

 amount of hydrogen, and is fixed in the slate of diamond. It is 

 known that hydrogen is present in the diamond, and also in steel, 

 and it is submitted that it forms the active agent in reducing the 

 carbon from the amorphous to the crystalline form. On ana- 

 lysing this hardened steel, it is supposed that the crystallii e 

 carbon goes off in all cases as gas ; so that less "combined 

 carbon will remain to be shown by the colour tests or deposited on 

 solution in hydrochloric acid. It must follow from this view that 

 carbon is the acting hardener of steel, and that the idea of other 

 elements, as phosphorus hardening steel is a delusion. In sup- 

 port of this it is observed that phosphorus does not harden 

 wrought iron and that probably the real effect of phosphorus 

 and silicen is to cause dissociation of carbon, thus producing a 

 larger extent of crystalli-ation and a harder metal. Thus it is 

 found that the higher the proportion of phosphorus, the greater 

 will be the difference between the carbin, as shown by the colour 

 test, and as fixed by analysis. Again, English Bessemer or 

 Siemens steel will require 20 per cent, less carbon to make it 

 work and harden equally veil with best Swedi-h steel; the 

 explanation being that the phosphorus in the former assists the 

 dissociation and crystallisation. To this effect of phosphorus 

 many of the mysterious failures of steel may probably be traced. 

 With regard to No. 3, the author regards the variations of 

 tempering as due solely to the completeness, or otherwise, of the 

 decomposition of the crystalline carbon in the hardened steel. 

 He observes that carbon and iron have no action on each other 

 at the heat at which tempering is effected ; while, even at such 

 temperatures, the abstraction of hydrogen from carbon, in the 

 presence of iron, cannot be deemed impossible. The tempering 

 of steel by simply quenching it in hot water or oil, may thus be 

 explained ; the outer layers may be supposed to be hardened at 

 first in the ordinary way, but then, as the interior heat does not 

 pass away si rapidly, it has time to act on the crystalline carbon, 

 and partly to dissociate it agai", thus producing something be- 

 tween hardened and unhardened steel — in other words, tempered 

 steel. The crystallised carbon in the hardened steel is supposed 

 to be diffused in a state of molecular disaggregation, and tobe 

 less intimately united with the iron than before hardening. 



As to No. 4, the startling statement that the ultimate strength 

 of steel is very little dependent on its amount of carbon, is ex- 

 plained to refer to the strength as calculated upon the fractured 

 area, n \ the original area. It is, therefore, equivalent to saying 

 that the contraction of the fractured area in iron or steel is pro 

 portional to the diameter of ultimate strength. The author finds 

 that this is the ca*c, both in the various pmblisbed tables of ten- 

 sile strength of steel, and in his own experiments. Hence be 

 holds that the contraction of area should be taken as the proper 

 measure of ductility (as is usual on the Continent), and not the 

 elongation. He looks upon hard steel as a metal of a certain 

 strength, having diffused through its mass a greater or less number 

 of particles of a very hard and rigid substance. Hence, as 

 ductility means the power of contracting in area, and extending 



in length by molecular flow, the ductility will be less as flow is 

 more difficult ; and flow will be more difficult as there are more 

 of the rigid crystals in the mass. The apparent strength per 

 unit of original area is thus increased ; but the strength per unit 

 of fractured area is usually diminishei, probably because the 

 hard sharp cryrtals tend to cut the metal between them, and 

 produce a sort of tearing action. For these reasons the use of 

 ductile and mild steel, in structures of every kind, is much to be 

 preferred to that of a brittle material, though of a higher 

 apparent tenaci y. 



A CHAPTER IN THE HISTOR Y OF CONIFER A? 



The Abietine/E 

 '"F HE most recent classification of the Abietineer, and 

 ■^ the one that will probably be chiefly adhered to, at 

 least in England, is published in the "Genera Plantarum " 

 of Bentham and Hooker, 18S0. In it Finns, Cedrus, 

 Picea, Tsuga, Pse'udotsi/oa, Abies, and Larix, are recog- 

 nised as separate genera. The tribe comprises the cedars, 

 larches, firs, pines, and contains some 150 species, and is 

 almost exclusively confined to northern and north tem- 

 perate regions. The genera are all cone-bearing, and 

 with few exceptions produce winged samaroid seeds. No 

 definite remains are known of earlier age than Jurassic, 

 but with the Wealden and Cretaceous they become 

 plentiful, and already in the Neocomian and Gault the 

 ancestors of several existing genera were completely 

 differentiated. 



Phius, Linn. — The cones in this genus vary from the 

 size of a walnut to a length of 19 inches, or possibly even 

 more. The scales are woody and persistent, and closed 

 until the seeds are ripened, when they gape widely. The 

 seeds are in pairs under each scale, and, with few excep- 

 tions, winged. The leaves are acicular, and in some cases 

 very long, and are sheathed in bundles of two, three, or 

 five. Nearly all classifications are mainly founded on the 

 number of leaves that occur in a fasciculus, but this 

 character is rejected in the " Genera Plantarum" as in- 

 constant. Two natural divisions are, however admitted — 

 Pinaster and Strobus. 



The former and larger division is distinguished by 

 the scales being very closely adpressed before shedding 

 the seeds, and by their quadrate, umbonate, or elongate, 

 conical heads. The Strobus section is comparatively 

 small, and has elongated, often pointed cones, with hard 

 and rigid, yet scarcely woody, loosely imbricated scales, 

 thicker centrally than at the margins, and terminating in 

 a minute or obsolete umbo. Cones of P. strobus and P. 

 excelsa, representing this section, may be picked up in 

 most botanical gardens, while the Pinaster section com- 

 prises all the pines commonly grown in plantations. 



Besides the " Genera Plantarum," many excellent 

 accounts of the tribe have recently been published. 

 Among these are Gordon's " Pinetum " (1880), Veitch's 

 " Manual of the Coniferas" (1881), Dr. Maxwell Masters' 

 "Conifenc of Japan " (Linn. Trans. 1881), and an ex- 

 quisitely illustrated essay on the " Coniferous Forests of 

 the Sierra Nevada,'' in Seribner's Magazine, also in 

 1881. 



Of the Pinaster division seventy-seven fossil specie-- 

 were enumerated by Schimper ; none, however, are 

 definitely assigned to the group from deposits older than 

 the Eocene of Aix, and most are from the upper Miocene, 

 and even later deposits. The oldest forms are from 

 Solenhofen, and the Gault of Hainault is said to contain 

 connecting-links between the two sections. 



Of the Strobus division twenty species are enumerated, 

 the oldest being from the Komeschichten of Greenland ; 

 but there are a number of additional species which can- 

 not well be grouped in either section. 



In England no cones are known that can be referred to 

 Pinus, as now restricted, from rocks older than the Pur- 

 beck, but their number gradually increases until the close 

 of the Tertiaries. 



