Hoba (South-West Africa), the largest known meteorite
By L. J. Spencer, MA, SC. D, FRS
In 1910 in volume 15 of this magazine I was able to place on record some details respecting the ‘Cullinan’ diamond from South Africa, the largest crystal of diamond yet found. I have now been able to collect some information about the largest known meteorite. The facts concerning remarkable objects such as these are apt to become exaggerated and distorted, and it is well to place on the record the true and accurate data while they can still be ascertained.
The scanty accounts so far published of the Hoba meteorite give the date of discovery as ‘about twenty’ or ‘about ten years ago’. Its weight has been variously estimated at 40, 50, 60, 70, 87, and 100 tons, and the dimensions mentioned have ranged up to 8 metres.
The locality (fig. 1) where the mass still lies is close to the southern boundary of the Hoba-West farm (no. 322) in the Grootfontein district. It is 12 miles (by road 15 miles) west of the town of Grootfontein (meaning in Dutch, great spring). The mass has consequently, not inappropriately, been referred to as the Grootfontein meteorite. Its position is approximately by latitude 19deg 35’ S, longitude 17deg 56’E. It lies a couple of hundred yards on the east side from the rough side-road leading to Otjihaenene siding on the railway from Grootfontein to Otavi, and only four kilometres from the narrow 2-foot gauge railway line.
On the German map (1:400,000) issued on 1910 by the Survey Department in Berlin, and reproduced by the English Ordnance Survey in 1915, this farm is marked as Paviansfontein, but in the estate map of the South West Africa Company it appears as Hoba-West. The Afrikaans form is Hoba-Wes. The adjoining farm to the east is known as Hoba-Ost (the German form for Hoba-East), and the homesteads of both farms form a group of substantial buildings where the main road crosses the farm boundary. On some maps this is marked simply as Hoba.
The exact date of the discovery of the mass I have not yet been able to ascertain, but it appears to have been late in 1920. Mr Michael H Hanssen the then occupier of the farm has since left. Dr Paul Range, formerly Government Geologist of German South-West Africa, who made a special study of the meteorites of the country, informs me that the Hoba meteorite was certainly not known previous to the war of 1914-18. The earliest mention of the mass appears to be in a letter dated May 26, 1921, from the late Mr T Tonnesen, General Manager in Grootfontein of the South West Africa Company, to the London office, which I am permitted to quote. Accompanying this letter there is a photograph of the meteorite partly exposed in the pit dug around it:
“On the farm Hoba-West, about 20 kilometres to the west of Grootfontein, there is a meteorite, which is about 2.5 by 2.5 metres on the surface and may have a thickness of probably 2 metres (only 1 metre has been exposed). Accordingly the weight would be 87 tons, but it is safer to say 60 tons. The assay shows 81.29% iron and 17.49% nickel. The nickel content would probably amount to 10 tons. The stuff is very ductile, precluding blasting. We shall try a saw and eventually an oxy-acetylene flame.”
Since all meteorites in South-West Africa are the property of the Government, nothing further was done in the way of mining for nickel.
In September 1921, Mr M H Hanssen wrote from Hoba-West to the Director of Geological Survey of the Union of South Africa in Pretoria, giving the dimensions of the meteorite as over 3 yards by 3 yards by 39 inches thick, and quoting the results of the assay made by the South West Africa Company.
The first published mention of the meteorite appears to be in Dr G T Prior’s “Catalogue of Meteorites” (British Museum, 1923, p73). This was based on a verbal communication from Prof Charles Palache, who visited the locality when on the Shaler Memorial Expedition from Harvard University to South Africa in 1922. [figure included] Some further information in the “Appendix to the Catalogue of Meteorites” (British Museum, 1927, p23) was supplied in 1926 by Dr A W Rogers, the Director of the Geological Survey of the Union of South Africa. This was based on the letter of Mr Hanssen mentioned above.
The next mention of the meteorite that I have been able to find is an anonymous note headed ‘the largest meteors’ in ‘Science’ of March 22, 1929 (New York, vol 69, p xii) which was reprinted in the American Mineralogist, vol 14, p 201:
“A seventy-ton meteor [ie meteorite] is reported to have been found at Otjihaene [ie Otjihaenene], near the head of the Grootfontein railway in the north-eastern part of Southwest Africa, imbedded in soft limestone. Its approximate size is ten by ten by four feet. Though this is said to be the largest actually discovered in the world, it is probably dwarfed by the one which many years ago caused the famous Meteor Crater in Arizona.”
In a special number of the “Zeitschrift fur praktische Geologie” issued in June 1929 on the occasion of the XVth session of the International Geological Congress in South Africa, Prof H Schneiderhohn gave a brief account of the Hoba meteorite. This was based on information supplied by Mr A Stahl and is accompanied by a photograph taken by Mr F W Kegel, the General Manager of the mining company at Tsumeb. He states that it was known since about 1920, and gives dimensions as 3 x 1.5 x 1.5 metres, and the estimated weight as 50 tons. Prof Schneiderhohn’s work in this district during the years 1914 – 1919 is well known, and in collaboration with Mr Stahl, who worked on the ground during 1922-26, a geological map has been made. But it was not until 1925 that Mr Stahl saw the meteorite. Incidentally, Prof Schneiderhohn gives a graphic description of a brilliant meteor (fireball) which he himself saw at Tsumeb on September 16, 1917; though this, of course, has no connection with the Hoba meteorite.
At the British Association meeting in South Africa in August 1929, a paper on “The New Grootfontein Meteorite” was read by Prof W J Luyten, Superintendent of the Boyden station at Bloemfontein of the Harvard University Observatory, who visited the meteorite early in 1929. This paper gives a quite inadequate description of the meteorite, and moreover contains some inaccuracies. The specific gravity is given as 7.96. Since then the meteorite was visited by myself in September 1929, and by Mr S G Gordon of Philadelphia in December 1929. A brief mention of it, as one of the ‘sights’ of the Grootfontein district, has recently been made by Mr A W Clark of the South West Africa Company. Lastly, a brief account of the meteorite has been given by Dr Paul Range, who visited it in September 1929, together with a description by Prof R Schreiter of the ‘iron-shale’ surrounding the mass.
The very successful excursion to South-West Africa organized by the International Geological Congress in 1929 ended on September 5 with a visit from Tsumeb to the Abenab vanadium mine and the large meteorite near Grootfontein. Here the party (a small one, as one motor-car broke down on the way, and the majority of the members had elected to visit the game reserve round the Etosha Pan) was conducted by Mr W R Feldtmann, the General Manager of the South West Africa Company. Unfortunately the visit was rather a hurried one and the time was all too short for any detailed study. Nor was it possible to collect any specimens beyond the ‘iron-shale’ surrounding the mass. The huge block of metal is roughly rectangular in shape with few irregularities and no projecting portions.
When first found, only a small portion of the meteorite was showing on the surface. A pit has since been dug partly round the mass; and at the one end, where the pit is about six feet in depth, a small portion of the under surface has been exposed. The top surface is approximately flat (fig. 2) and level with the surrounding ground. A dozen people can walk round on the level surface. No attempt has been made to move the mass, and it is not known whether the under surface is flat or not. The mass is embedded in a white, soft and porous, surface limestone (Kalahari Kalk) from which it is separated by a layer one foot thick in thickness of laminated ‘iron-shale’ (figs. 3 and 4). This ‘iron-shale’ is dark-brown to black in colour with a dark-brown streak, and is magnetic. It consists largely of limonite with some magnetite and perhaps also trevorite (Ni Fe2 O4). It shows green nickel stains and is seamed with white calcium carbonate from the surface limestone. The lamination of the shale follows the contour of the meteorite, being vertical at the sides (fig. 3) and horizontal at the base (fig. 4). There is a sharp separation between the shale and the metal, though the metal shows a little rusting and scaling on its surface. The rock underlying the surface limestone at this spot is granite, close to the junction with the dolomite of the Otavi formation. The ground is level and there is no crater around the meteorite.
The dimensions of the block taken with a steel tape are 295 x 284 cm (about 10 x 9 feet) on the flat surface, with a thickness at one end of 122 – 111 cm, and at the other of 75 – 55 cm. With a specific gravity of 7.96 this corresponds to a weight of 60 metric tons for the block. The block shows a complex of broad and shallow concave surfaces. There are no large and prominent pits and no angular corners. It is just a huge block from which it is difficult to detach any portion. I have previously suggested that this characteristic surface of iron meteorites is the result of slow atmospheric weathering, rather than the result of rapid burning during the brief flight through the earth’s atmosphere. The foot of ‘iron-shale’ seen in situ represents the amount of weathering that the mass has undergone since it has lain in the ground. This perhaps indicates the original size of the mass; adding 30cm all round, an original weight of 88 tons would be indicated (curiously the same as Mr Tonnesen’s first estimate).
Since my visit, several further visits have beeb made by Mr Feldtmann, and he has very kindly supplied me with some further information, including the exact position of the meteorite, as indicated on the sketch-map (fig. 1). With a magnetic compass he found that the mass is magnetised with polarity. The (geographically) north end of the block attracts the south end of the needle and vice versa, and along the south is a median neutral zone. For the purpose of determining the magnetic axis of the meteorite a six-foot length of steel drill was magnetised, and Mr A W Clark found that the axis is about 14deg east of the earth’s present magnetic axis. No doubt the meteorite became magnetised in the earth’s field when it fell with a violent concussion; but whether it would be possible to deduce from this figure the date of the fall is very doubtful.
Mr Feldtmann has also been able to secure a piece of the meteorite, which was generously presented in 1930 to the British Museum by the South West Africa Company. After obtaining official permission from the Administrator of South-West Africa and from the Department of Mines, the piece was sawn off under his personal supervision. This was done by two natives working for two days and with a considerable consumption of hack-saw blades. The specimen (BM 1930,976) obtained is a wedge-shaped piece weighing 2489 grams (5.5 lbs) with a triangular cut surface measuring 22 x 9 cm (about 15 square inches). It shows two earlier cuts that had been made surreptitiously by unauthorised persons, and also the willful damage done with an oxy-acetylene blowpipe. After smoothing and polishing the smallest of the three cut surfaces, and cutting off a piece from the thin end for more convenient microscopical examination, the specimen now weighs 2444 grams. Mt Feldtmann also sent 83 grams of the sawdust, which is still available for any further investigation.
The metal is comparatively soft and quite malleable, and is more easily worked than some other meteoritic irons. It takes a brilliant steel-grey polish. The larger polished surface (7 x 6 cm) shows a few (five) rounded or elongated bronze-coloured areas of troilite, the largest 3 x 0.5 mm, and the others much smaller. Also numerous minute tin-white specs of schreibersite (or cohenite?). Both the troilite and the schreibersite (?) are softer than the nickel-iron groundmass, and they are worn down to a lower level. From each speck of schreibersite (?) there is a minute pit with the shape of an acute isosceles triangle, the speck being at the narrow base-line. These pits are parallel over the whole surface and they all point in the same direction. From their sides, first one and then the other, they reflect light simultaneously over the whole surface. The effect was, however, produced only after the surface had been much worn by repeated polishing.
Etching the polished surface with very dilute nitric acid for a few minutes reveals no distinctive structure. Even with longer etching in stronger acid there are no indications of Widmanstatten figures or Neumann lines. The surface is merely dulled. At a certain angle, however, a reflected sheen is shown uniformly over the whole surface, except on a few scattered areas of small size. At another angle the latter show a reflected sheen all together, and the main area is then dull (fig. 5). This effect is mentioned and figured in S G Gordon’s paper (loc. cit. 1931, p 253). The main area evidently represents a single crystal individual, and the smaller portions are no doubt in twinned position with respect to the main portion. The boundaries between the two areas are sharply defined, sometimes with straight edges and sometimes dovetailed into each other, suggesting lamellar twinning.
The minute tin-white specks of schreibersite (?) show up more prominently on the etched surface. They are of irregular distribution and have the form of slender needles or curved hairs. In cross-section they show as minute circular areas, scattered singly over the surface or aggregated into small clusters. With rather deeper etching the needles and hairs show a groove along the axis, and the circular areas show a sunken centre.
The larger polished surface on the main specimen is adjacent to the edge that has been fused with an oxy-acetylene blowpipe. The very thin fused layer was not attacked by the etching reagent, and it forms a narrow (0.1mm) bright border on the etched surface. Beyond this, the heating appears to have had no effect on the iron.
A second polished surface (3 x 1 cm, fig. 5) on the small piece cut from the thin end of the specimen and 18cm away from the fused area, has been kindly examined by Dr J M Robertson at the Royal School of Mines, London, with the permission of Prof Sir Harold Carpenter, FRS. The section was repolished and very lightly etched with a 1% solution of nitric acid in alcohol. His photomicrographs (plate I) taken with an oil-immersion objective show under a magnification of 820 diameters a minute duplex structure somewhat similar to that of plessite. Fig. 7, taken at the junction of the two portions that show oriented sheens at different angles, shows this structure with two directions. Figs. 10 and 11 of one of the schreibersite (?) specks show a border of pale yellow material with a white inner portion. With further etching, as noted above, it appears to be this central portion that is more deeply attacked.
The Hoba meteorite is to be classed as a nickel-rich ataxite. Previously known meteoritic irons of this class are tabulated below in the order of their nickel percentages. None of these nickel-irons was observed to fall, and the supposed meteoric origin of the last two with exceptionally high content of nickel has been questioned. In view of its present situation, any origin other than meteoritic for the Hoba mass is inconceivable.
The Mineralogical Magazine/The Journal of the Mineralogical Society
N0 136, March 1932, Vol XXIII
Update: Nov 2013
I’ve been contacted by Professor Peter Spargo (University of Capetown) who has kindly supplied me with the following articles. They were first published in the MNASSA (monthly notes of the astronomical society of South Africa).
The Hoba Meteorite: Part 1
The Hoba Meteorite: Part 2
The Hoba Meteorite: Part 3