FEBEUAEt 1, 1899.] 



KNOWLEDGE. 



33 



tion other more or less authenticated instances. In 1852, 

 Prof. Andrews," of Btlfast, powdered np certain Irish 

 dolerites and basalts in a porcelain mortar, extracted all 

 magnetic particles by means of a magnet, and examined 

 these particles under a microscope. When moistened 

 with an acid solution of copper sulphate, the magnetic 

 iron oxide (magnetite) undergoes no change ; but metallic 

 iron is attacked, and is replaced by a deposit of pure 

 copper. Andrews found in this way a small number of 

 particles of pure iron — three or four from one hundred 

 grains of powdered rock — and the largest deposit of copper 

 measured one-fiftieth of an inch in diameter. The dolerite 

 of Slemish yielded the best results ; but the basalt of the 

 Giant's Causeway gave suiBcient indications. 



Other workers repeated these observations on Con- 

 tinental rocks, the result being that metallic iron seems 

 regularly associated with the dark heavy " basic " rocks, 

 basalts, and so forth, and to occur also, but more rarely 

 in rocks rich in silica, such as granite. 



But it is only in the last few years that geologists have 

 been assured of the presence of considerable masses of 

 native iron in the earth's crust. Nickel-iron has long 

 been used by the Eskimo for knife-blades and hatchets ; 

 and it was suspected that some large meteorite was being 

 worked as a convenient mine. Nordenskiiild, in 1870, 

 observed a block of iron in some ballast taken on board 

 near Godhavn, in Disko Island ; and in time the true 

 source was discovered by him at Uigfak, or Ovifak, a 

 spot difSnult of access, but now one of the best known 

 mineral localities in the world. Here, on the shore, some 

 magnificent blocks of nickel-iron were lying, exposed to 

 the sea-waves. The largest, which is now in Stockholm, 

 measures six feet long, and weighs about nineteen tons. 

 On etching, the well-known " Widmannstiitten figures " 

 were revealed, and nickel, varying from 1 to 6'5 per cent, 

 (usually about two per cent.) has been found on chemical 

 analysis. 



It is no wonder, then, that the Ovifak masses were at 

 first believed to be meteoric. Even when nickel-iron was 

 found in the adjacent basalt of the coast, it was thought 

 that a meteoric shower might have occurred while the 

 lava was still hot and viscid, and that the fragments had 

 become consequently entombed. But other discoveries 

 followed in the same neighbourhood, notably those of 

 Steenstrup, in 1875, and it became admitted that the iron 

 was a true constituent of the Miocene basaltic lavas. 

 Several writers then saw in it the result of the reduction 

 of iron oxides by the material of a coaly bed, which the 

 lava was supposed to have absorbed ; so reluctant were 

 some geologists still to admit that native iron is, in any 

 quantity, a true constituent of the earth's crust. The 

 carbon present in the nickel-iron, amounting often to two 

 per cent., might be attributed to the same source ; but 

 traces of carbon are well known to occur in most truly 

 meteoric irons. Dr. Lawrence Smith showed that metallic 

 iron was included in even the felspar of the Disko basalts, 

 and the meteoric theory, at any rate, may be said to have 

 been abandoned for fully twenty years. 



Meanwhile, attention was called to other examples of 

 terrestrial iron, commonly alloyed with nickel, and some- 

 times associated with magnetic pyrites (pyrrhotine). Let 

 us note in passing that troilite, the iron sulphide frequently 

 found in iron meteorites, is probably identical with the 

 pyrrhotme of our own earth's crust. Tuis later series of 

 discoveries only confirms the belief that terrestrial iron is 

 typically associated with the most basic masses of the 



• Seport Brit. Assoc, 1852. p. 34. See also " Scientific Papers of 

 Thomas Andrews " (1889), pp. 231, 399, and 507. 



crust. These basic masses have been, on many grounds, 

 regarded as more deep-seated in their origin than the 

 lighter granites and gneisses, which play so large a part 

 in the constitution of the crust. 



Near Biella, in Piedmont, grains of nickel-iron occur in 

 a gold-bearing sand ; here the nickel forms seventy-five, 

 and the iron twenty-five per cent. In Oregon, pebbles 

 have been found containing twenty-three per cent, of iron, 

 with sixty per cent, of nickel, and twelve of silica. The 

 most remarkable discovery of this nature, however, is 

 probably the " Awaruite " of New Zealand. In 1885, 

 Mr. W. Skey observed this substance as small grains in 

 a sand from the Gorge River, near Awarua (Big Bay), on 

 the west coast of the South Island of New Zealand. Its 

 composition shows that it is a nickel-iron, in which there 

 are two parts of nickel to every one of iron. Prof. Ulrich 

 caused its true source to be traced in the wild country 

 north-east of Awarua ; and it is found to occur with pyrite 

 and chrome iron ore, scattered through typical peridotites. 

 The "peridotites" are rocks rich in magnesia and iron 

 oxide, and containing only some forty-two per cent, of 

 silica ; they have been recognised, all the world over, as 

 the natural matrix of ores of chromium and nickel. 

 Metallic platinum is also traced to peridotites ; and metallic 

 iron IS recorded, as we have seen already, from the platini- 

 ferous sands of Brazil. 



Here we have a series of links, which make it now 

 obvious that ma- 

 terials, once supposed 

 to be entirely 

 meteoric, may occur 

 in mass in the lower 

 regions of the crust. 

 Mr. Skey compared 

 his "awaruite" with 

 the substance found 

 at Octibbeha City, 

 Mississippi, in 1857, 

 which has been 

 already referred to as 

 contair:ing some sixty 

 per cent, of nickel. 

 He was criticised for 

 not observing that 

 the " octibbehite " 

 was of celestial 

 origin ; but the tables 

 have now been turned 

 by the suggestion 

 that both this and the 

 nickel-iron of Santa 

 Catharina are not 



meteorites, but have been washed from some portion of our 

 own crust. How many of the great iron masses found in 

 spots difficult of access, and hitherto regarded as meteorites, 

 may ultimately be traced to some igneous mass, iutruded 

 through the local strata '? 



Already the abundant specimens of " meteoric " iron 

 from Cdiion Diablo, Arizona, have come under suspicion. 

 There is now, in fact, no certain character left to us by 

 which we can decide whether a nickel-iron mass is of 

 celestial or of terrestrial origin, except, perhaps, the nature 

 of the surface. This presents remarkable pittings and 

 depressions, in the case of iron that has been projected 

 with great ve'ocity through the air.* 



* See on this matter and for an examination of meteorites in 

 general, the classic work of Daubreo " Etudes sjntlietiques de 

 geologie eiperimentale " (1879), pp. 473-703. 



Fia. 2. — Cut eurface of Basalt from 

 Ovifak, Disko Id , Oreeuland. x 12. 

 Bright irregular specks, composed al- 

 nost entirely of nickel i'on, lie 

 scattered in a riark ground, wliich 

 consists of the silicates of the basalt. 

 Compare with Fig. 1. ( From a speci- 

 men m the Museum of Science and 

 Art, Dublin.) 



