The origin of mineral veins, including those in which gold is contained, has long been discussed by geologists. The old theory that the quartz of veins was originally in a molten condition and was ejected from below into fissures is no longer maintained, although in 1860 H. Rosales brought forward evidence in its favour as far as the Victorian lodes are concerned. One of the theories now strongly advocated is that the materials forming the veins have been transported in aqueous solution and precipitated where they occur. In a few exceptional cases, sublimation may have played a part. The view that the solutions found their way downwards from above has been abandoned, but the ascensional theory and the lateral secretion theory both have many adherents. The last-named theory found its principal supporter during many years in Prof. F. von Sandberger, who pointed out that the gangue of many lodes varies in composition if the nature of the rocks through which they pass is changed, and claimed to have proved by analysis that the materials forming vein-stone are derived from the adjacent country rocks. He stated, moreover, that such minerals as augite, hornblende, mica, and olivine, which are essential constituents of crystalline rocks, contain small quantities of the heavy metals occurring in veins. Although Sandberger did not try to detect gold in the silicates, this metal is not likely to be an exception. Prof. A. Stelzner objected to these conclusions, urging that small quantities of the sulphides of the heavy metals were probably mechanically mixed with the crystals of minerals which Sandberger analysed in the belief that they were pure. Stelzner advocated the retention of the ascensional theory which alone affords a satisfactory explanation of the difference in composition observable in neighbouring lodes passing through the same rocks, and apparently formed at different periods. The two theories are, however, not contradictory, and perhaps neither need be entirely rejected, the solutions being supposed to pass more or less freely in the plane of the lode after they have been impregnated. 
In 1897, Dr. Don announced, as the result of several years work in Australia and New Zealand, that the gold in those regions if present in country rock is invariably contained in pyrites. He found that the amount of pyrites and its richness in gold diminished with increase of distance from an auriferous lode, and the gold soon disappeared. In the various minerals of igneous rocks, augite, hornblende, mica, &c., no gold could be detected, although by his method the limit of accuracy was 0.05 grain per ton. He concluded, therefore, that the gold in these lodes was not derived from the country rock adjacent, but that the latter was impregnated with gold from the lode by solutions rising from a depth greater than that of any of the rocks exposed at the surface. L. Wagoner doubts the accuracy of these statements of Dr. Don, finding that they do not agree with his results when examining the rocks of Western America. He finds gold in rocks apparently free from sulphides.
J. E. Spurr traces the origin of certain quartz veins direct from igneous magmas by a process of magmatic segregation, and derives an auriferous quartz vein from an igneous siliceous dyke. The magmatic concentration is succeeded by aqueous concentration acting underground by dissolution and precipitation, and on the surface chiefly by mechanical means, in both cases producing ore bodies.
How Gold is Formed
- From veins in rock formations. In this division may be included replacement. deposits, disseminations in rocks, and also, for example, the marine deposits accumulated in shallow water, such as the conglomerates of the Transvaal.
- From placers or the alluvial deposits of ancient and modern streams. Modern beach deposits and loose sands or gravels generally, may be included in this section.
One of the most striking differences between the ores of gold and those of all other metals lies in the extremely small proportion which the desired material bears to the worthless gangue with which it is accompanied. Occasionally hand specimens in vein stuff are found containing several per cent, of the precious metal, but these are of quite exceptional occurrence and have not the slightest economic importance. The greater part of the vein gold now being produced is derived from ores containing only about one part of gold in seventy or eighty thousand, whilst, under exceptional circumstances, a yield of one part in half a million parts of gangue may give handsome profits. Placer deposits are usually much less rich than this; the average amount of gold contained in those now worked does not exceed one part in one million, and in California deposits of gravel with only one part of gold in fifteen millions have proved susceptible of successful treatment by hydraulic mining on a large scale.
- Vein Gold: In this case the metal, whenever it is present in visible grains or masses, has sharp angular edges, although it is usually not distinctly crystalline. It frequently penetrates the rock irregularly in various directions, and is completely interwoven with, and attached to the matrix, usually quartz, so that the metal cannot be separated from the rock without crushing the latter.
The gold in lodes is sometimes in the form of crystallisations, which are, however, exceedingly rare, and crystals of gold are still probably unknown to most miners, although they occur more frequently in placer deposits. Arborescent branching and dendritic masses of crystalline gold are more common than single crystals in both quartz lodes and placer deposits. The crystalline forms met with have already been described, p. 7. In the Transylvanian lodes, gold occurs chiefly in thin sheets or plates, often as much as from half an inch to two or more inches in breadth. Such plates are rarely thicker than a visiting card, and are generally covered with crystalline lines and markings, revealing a distinct geometrical structure. Gold also occurs in wire-like forms, sometimes penetrating crystals of other minerals, such as calcite and dolomite.
It frequently happens that the gold in lodes, &c., is in a state of fine division and is not visible without magnification.
The matrix in which the gold is contained is usually quartz, but occasionally gypsum or other minerals, intersecting as veins or interlaminated with sub-crystalline, slaty, or schistose rocks, especially hydromica and chloritic slates. Gold also occurs sparingly in similar veins in granite and gneiss. It also occurs in the midst of rock formations without any obvious connection with quartz veins. Among such rocks are granite, aplite, gneiss, eurite, quartz-trachyte, syenite, andesite, basalt, diorite, gabbro, diabase, schists, porphyry and slate. These deposits seldom pay to work and in that case can hardly be called gold ores.
J. E. Spurr writes—“Native gold has been found in both basic and acid rocks. It has also been detected in the dark ferro-magnesian silicates of rocks of all degrees of acidity. The commercially-valuable concentrations of gold are generally connected, now with basalt or gabbro, now with diorite, now with phonolite, rhyolite or granite. They occur also in many different forms, as replacement-deposits in limestone, as disseminations in igneous and sedimentary rocks, as contact-deposits near intrusive masses, and in fissure-veins.”
- Placer Gold and Nuggets: Placer gold is usually in the form of small scales, but pellets or rounded grains also occur, while large masses or nuggets are usually of a rounded mamillated form. The chief difference between the appearance of placer and vein gold lies in the fact that the former is always rounded, showing no sharp edges, even the crystals having their angles smoothed and rounded off. This has been pointed to by the advocates of the erosion theory of the origin of placer gold, as evidence in favour of their views, the roundness of the fragments being taken to prove that abrasion of the gold has been affected by attrition with water and grains of sand. The largest masses of gold yet discovered have been found in auriferous gravel. The “ Blanch Barkley ” nugget, found in South Australia. weighed 146 pounds, and only 6 ounces of it were gangue; and one still larger, the “Welcome” nugget, from Victoria, weighed 2,195 ounces, or 183 pounds, and yielded gold to the value of £8,376 I Os. 6d. In Russia a mass was found in 1842 near Miask, weighing 96 pounds troy. The largest mass from California is given in the State Mineralogist report as weighing 2,340 ounces, or 195 pounds, but no authentic cases seem to be on record of nuggets from this State weighing more than 20 pounds.