conclusions. In this paper I will confine myself to a short account of the tin occurrences at Ben Lomond.*

The granite of the Ben Lomond district is composed principally of quartz and feldspar, with very little biotite. Much of the feldspar (orthoclase) occurs in large well-defined idiomorphic crystals of the Carlsbad twin type, and these are porphyretically distributed throughout the finer-grained ground-mass of the granite. At the Mt. Rex Mine a large tinstone deposit, some 80 feet in width, occurs. This stone is not a typical greissen, as the mica is only recognisable microscopically, but from a genetic point of view the stone is essentially a stanniferous greissen. It is evidently a metamorphosed form of the surrounding granite, for embedded in the greissen are found various pseudomorphs after feldspar, the shape of the Carlsbad twin crystal being distinctly recognisable. During my visit to the mine in June last year I was able to identify the outline of the porphyritic crystals of feldspar in the following minerals:-(1) cassiterite, (2) cassiterite and quartz, (3) confused muscovite and calcite, (4) chlorite, (5) chlorite and quartz, (6) tourmaline (the latter from the section north and adjoining the Mt. Rex Mine). As far as I am aware, the only other locality where cassiterite has been found pseudomorphous after feldspar is at Wheal Coates, near St. Agnes, Cornwall. They were first identified in the Mt. Rex deposit by Mr. W. F. Petterd, of Launceston. They are not so dense nor so clean cut as some of the specimens from Cornwall, but they are unmistakably pseudomorphs. The occurrence of pseudomorphs amounts to a definite proof that all these minerals may occur as a replacement of feldspar. The deposit is traversed by veins of quartz and fluorspar, but no topaz has yet been found in the district. Quartz veins are abundant both in the granite and the overlying Silurian slates, and the latter may often be shown to pass over into veins of pegmatite. In a great many instances I found feldspar with the quartz, and specimens could readily be collected representing every intermediate stage between the pure quartz vein and the typical coarsely crystalline quartz-feldspar-mica pegmatites; the latter, however, are much less common than the quartz veins of the intermediate types. Both the quartz veins and the pegma

*The following reports on tin mining districts in Tasmania have been issued by the Mines Department of Tasmania since the above was written :-W. H. Twelvetrees, F.G.S., Government Geologist, Report on the Tin Mines of the Blue Tier, County of Dorset; G. A. Waller, Assistant Government Geologist, The Tin Ore Deposits of North Dundas; G. A. Waller, The Tin Ore Deposits of Mount Heemskirk.

tites sometimes contain tin oxide, with or without tourmaline, and there is one instance of an extraordinarily rich patch of tin oxide occurring in a pegmatite vein (the old Lomond Mine). Another stanniferous pegmatite vein contains giant crystals of beryl, which Mr. W. H. Twelvetrees has shown by microscopic examination to contain numerous fluid enclosures of carbonic acid, with moving bubbles. The pegmatite veins are also often accompanied by a certain amount of greissenisation, meaning by the latter term not the formation of typical greissen, but the modified form of that rock, which is common throughout the district. It may be mentioned that true stanniferous greissen occurs at the Roy's Hill Tin Mine, a granite contact deposit some 15 miles from the Ben Lomond district proper.

The age of the tin deposits in the Ben Lomond district, though it has not yet been so accurately determined as in some other districts, is evidently nearly as great as that of the granite itself. The latter is younger than the Silurian strata in which it forms intrusions and has produced metamorphism, and older than the Permo-Carboniferous strata which rest horizontally upon its eroded surface. It is therefore probably of Devonian age. The stanniferous quartz veins penetrate the Silurian strata, but are cut off sharply at the contact of the granite and the Permo-Carboniferous strata. The bottom layers of the latter are also often composed of an old stanniferous wash. This is believed to be the oldest known occurrence of stanniferous gravels in the world.

It is therefore evident that the tin veins were formed before the Permo-Carboniferous, and presumably during the Devonian Period. They belong, therefore, at least to the same geological period as the granite. The granite is traversed by numerous aplitic dykes, but up to the present no case has been recorded in which a tin vein has been intersected by one of these. The aplitic (quartz feldspar) dykes, however, sometimes contain considerable quantities of tin oxide, and even a little galena, both of which have the appearance of being original constituents of the rock.

As early as 1840-50 Elie de Beaumont and A. Daubrée sought to explain the formation of tin veins by pneumatolysis. They proved the stability of stannic fluoride at high temperatures, and assumed that the tin had ascended in this state of combination from the deep-seated granitic hearth," together with fluoride of boron, silicon, and gaseous chlorine, and phosphorous compounds. They supported this hypothesis by pointing to the prominent part which compounds of fluorine, boron, and phosphorous take in the

mineralogy of the characteristic tin-stone deposits. As further evidence, they furnished the results of exhaustive experiments in the production of tin oxide from volatile compounds. In these experiments stannic fluoride could not be used owing to difficulties in manipulation, but they were successfully carried out on the analogous compound stannic chloride. By introducing the vapours of stannic chloride and water into a white-hot porcelain tube, hydrochloric acid and small crystals of tin oxide were obtained. The reaction takes place according to the formula

Sn Cl, +2 H, O = 4 H Cl + Sn O2

We have no reason to doubt that the entirely analogous reaction

Sn F, 2H, O 4 H Cl + Sn O,

would take place under similar circumstances.

By similar experiments, with suitable vapours acting on each other or upon solid substances, many of the accompanying minerals of tin-stone veins were produced, including a substance of analogous composition to topaz.

Recent investigations of the phenomena of magmatic segregation and the formation of pegmatite veins have thrown much light upon the origin of tin veins. For it is now evident from the close connection which exists between tin veins and pegmatite veins, and from the fact that they have so many characteristic minerals in common, that they have been formed from solutions which have a common origin. Now, a similar connection to that which has been shown to exist between pegmatite veins and tin veins has also been proved to exist between pegmatite veins and aplite dykes. As bearing on this point, I might mention the presence of tin oxide in both the aplite dykes and the pegmatite veins at Ben Lomond. It seems, therefore, evident that if aplite dykes are of eruptive origin, so also are pegmatite veins, and after the exhaustive treatment which this question has received at the hands of J. Lehmann, W. C. Brögger, Iddings, and many others during the last few years, the plutonic origin of pegmatite veins is now, generally regarded as fully established. The differences between the pegmatite veins and aplite dykes may be satisfactorily accounted for, on the assumption that the pegmatites were deposited from a very much more aqueous magma than the aplites. It appears to be probable that at very high temperatures and pressure magma and magmatic water are miscible in all proportions. If this is the case, we may have. a perfect transition from molten magma to aqueous solutions on the one hand, and from eruptive dykes to mineral

veins on the other.

The series, granite-porphyry dykes, aplite dykes, pegmatite veins, tin veins, is in perfect agreement with this theoretical proposition.

The details of the process by which the magma becomes differentiated into sub-magmas containing varying amounts of water have given rise to much speculation, but there is reason to hope that very soon a sound working hypothesis will be evolved which will explain the division of the magma, not only into highly aqueous and slightly aqueous submagmas, but also into acid and basic sub-magmas. The subject is too complex to be discussed in this paper.

Whatever the determining causes are then, the granite magma is believed to separate itself into sub-magmas, containing differing amounts of water. Into the most aqueous of these, Arrhenius concludes, on chemico-physical grounds, that the ions of carbonic acid, hydrogen sulphide, chloric, fluoric, boric acids, silica, the alkali metals, Tess frequently the alkaline earths and the metals, iron, zinc, lead, copper, and tin would become concentrated. This process has been termed "acid extraction" by Vogt, because fluorine and chlorine are believed to play the principal part in effecting the concentration or extraction of the metals.

The aqueous portion is, of course, very small in comparison to the rest of the magma. Its amount depends upon the original amount of water in the magma, and no doubt in many cases there may not be sufficient water in the magma to produce the division at all. Where it is formed, however, the aqueous portion would be much more mobile than the other part, and would retain its fluid state at much lower temperatures. Owing to the consolidation of the granite, much of it would be expelled through cracks and fissures, and coming up into the cooler portions of the granite would first of all precipitate the normal constituents of the granite, the quartz, feldspar, and mica. The strong acids, still in an excessively hot condition, would attack minerals of the wall-rock, and a series of complex reactions would take place which, after the experimental results of Daubrée, we may well imagine would account for all the phenomena which we meet with in tin ore deposits. When dealing with phenomena which are so far removed from our observation as the processes of consolidation of deep-seated eruptive rocks, which take place at such high temperatures and pressure that they cannot be imitated in the laboratory, it is not surprising that our notions of these processes should be somewhat vague. This is certainly the case with our modern theories of magmatic differentiation and extraction. Nevertheless, although there may be much

that is crude and incomplete in the details of our theories, the main facts of magmatic differentiation and extraction may be regarded as scientifically proved. This is a great step in advance in the theory of ore deposits. For, in magmatic extraction we find an alternative process of concentration to that which has been so carefully investigated by the meteoric school. The adherents of the latter school have shown how concentration may take place after eruptive rocks have solidified. The plutonic school have shown how this may take place before or during consolidation. The working agent of the meteoric school is meteoric water; the working agent of the plutonic school is plutonic or magmatic water. According to some geologists, even magmatic water is of meteoric origin; but this question need not be discussed in the present paper. As far as the origin of ores is concerned, they must be regarded as quite separate.

Apatite Veins.

Professor J. H. L. Vogt* has made an exhaustive investigation of the apatite veins which occur abundantly in Norway, Sweden, and Canada, and he comes to the conclusion that these apatite veins bear precisely the same relation to basic rocks that tin veins bear to the acid rocks.

The following is a very free and somewhat abridged trans lation of his brilliant comparison of the two types of

veins:

1. Both the tin and the apatite veins represent fairly welldefined groups of ore deposits, both of them intimately connected with eruptive rocks: the tin veins with the acid eruptives granite and its dyke-forming and effusive representatives; the apatite veins with basic eruptives, such as gabbro.

2. The tin veins, and also the apatite veins, are certainly younger than the eruptive rocks with which they are connected, nevertheless, it may often be proved that the difference in age between the formation of the veins and the consolidation of the eruptive rocks is not great (the apatite veins are sometimes crossed by dykes of diabase, the dykeforming representative of the gabbro, in the same way that tin veins are crossed by aplite, the dyke-forming representative of the granite).

3. Still more characteristic for both groups of ore deposits is the remarkable metamorphism of the wall-rock, greissenisation in the case of tin veins, scapolitisation in the case

Zeitschrift fur practische Geologie, 1894-95. Beiträge zur classification der Erzvorkommen, &c.