June 27, 1895] 



NATURE 



21 



It is thus difficult to know the amount of elevation of these 

 rocks, Ijut about latitude 50° the base of the cretaceous must 

 in several places have considerably exceeded 10,000 feel in 

 altitude. 



Symo)iis Mottt/ily Mctcoroh^ical Magazine, June. — The 

 prmcipal article deals with rainfall in China, with remarks 

 by the editor, based on observations made from 1886-92, and 

 published in various places by Dr. Doherck, of Hong Kong. 

 The mean annual rainfall is small in the north, and increases 

 greatly towards the south. In the Gulf of Pe-chi-li the fall is 20 

 inches, but reaches double that amount in the Delta of the Vang- 

 Tse-Kiang, 58 inches at Hankow, and 68 inches at Ningpo. In 

 Formosa it ranges from 60 to 90 inches, but at Keelung, in the 

 north-east, it reaches 148 inches. The seasonal rainfall is 

 shown in tables divided into six districts. Notwithstanding the 

 proximity of most of the stations to the sea, the distribution is, 

 generally speaking, of that type which prevails over the greater 

 part of Asia. 



SOCIETIES AND ACADEMIES. 



London. 



Royal Society, January 24. -" Micro- Metallography of 

 Iron." Part I. By Thomas Andrews, F.R.S. 



In tile course of a research with high microscopical pou'ers 

 (including 300, 500, Soo, 1200, and upwards to 2000 diameters) 

 on the micro-crystalline structure of large masses of wrouglit 

 iron, the author observed the following novel metallurgical 

 facts : — 



When large masses, several tons in weight, of practically pure 

 wrought iron were allowed to slowly cool from a white heat, a 

 sec<jndary or subcrystallisation of the metallic iron occurred. 

 The n<jrmal primary crystals ot the iron, or those which have 

 hitherto been regarded as constituting the ultimate structure of 

 the metal, were found to enclose a subcrystalline formation con- 

 sisting of very minute, and much smaller, crystals of pure iron 

 also belonging to the regular order of crystallisation. These 

 crystals sometimes manifested the hexagonal form, the pre- 

 dominant angle being about I20', and often they assumed the 

 form of simple cubes. The secondary crystals were contained 

 within the area of the larger primary crystals. 



Typical illustrations of this duplex crystallisation found in two 

 large iron forgings are given in Figs. I and 2, and the relative 

 dimensions of a number of individual crystals are given in the 

 paper. 



The results of twenty measurements of the primary crystals 

 and twenty measurements ot the secondary crystals taken on each 

 forging are given on these tables. 



The markings of the intercrystalline spaces or junctions of 

 the secondary crystals were very clearly defined, but they were 

 exceedingly minute. The general form, contour, and relative 

 size of the priniary and secondary crystals, as seen in section, 

 will be noticed on reference to the accurate tracings. Figs. I and 

 2. The linear dimensions of the primary crystals would average 

 about O'oi inch, the linear dimensions of the secondary crystals 

 aver.aging about O'OOI inch. 



Judging roughly from the indications of the average micro- 

 measurements, there would appear to be approximately 

 1 ,000,000,000 of the secon<lary crystals in a cubic inch of the 

 metallic iron. 



In the case of both the primary and secondary crystals the pre- 

 dominant well-defined angles of the facets of the crystals hovered 

 more or less about the angle of 120°. The majority of the angle 

 readings, made with the goniometer attached to the microscope, 

 indicating generally a hexagonal structure on form of crystal- 

 lisation. There were, however, also perfect cubical crystals 

 observe<l. 



The observations were made with a Ross first-class microscope. 

 The micro-measurements afford an indication of the comparative 

 size of the primary and sect)ndary crystals. These measurements 

 were carefully taken by a Jackson micrometer, and in some 

 cases by a Kamsden screw micrometer, both accurately calibrated 

 with a standard stage micrometer. The wrought iron forgings 

 on which the observations were made were constituted of 

 practically jnire hammered wrought iron, the dimensions of the 

 mass being about 10 feet long and about 12 inches .square. The 

 great length of lime required for such large masses of iron to 

 i-ool from a white heat appeared to facilitate the production of 

 the crystals of the secondary formation. 



^■O. 1339. VOL. 52] 



The rationale of this duplex crystallisation has apparently been 

 as follows : — The mass of metallic iron on cooling having reacherl 

 the crystallising point at about 740' C. , the peripher)' or skelet<ms 

 of the larger or jirimary crystals were then formed. As the 

 period of cooling was, however, very slow, the semi-fluid or 

 viscous metal in the interior of these primary crystals was, on 

 finally consolidating, ajiparently further broken up or sutxtivided 



into a considerable number of smaller cr)'stals, enclosed within 

 the boundary or periphery of the primary crystals. 



In the course of further experiments on the cooling of large 

 masses of wrought iron, the author has also found, by the use of 

 high power objectives, that the secondary crystals sometimes 

 enclosed a still more minute form of crj-stals of pure iron, of the 

 cubical form, which may hence be regarded as constituting a 

 tertiary system of cr)-stallisalion in pure metallic iron. These 



experiments therefore indicate that large masses of heated 

 wrought iron, on cooling from above the temperature of the 

 crystallisation of metallic iron, viz. 740' C. , are capable of 

 crystallising in three distinct modifications which may tentatively 

 be called the primary, secondary, and tertiary system of crystal- 

 lisation in iron, these various crystalline modifications being all, 

 however, connected with the regular system of crystallisation. 



