If nickel iron (19-49 per cent.), troilite (14-43 per cent.), and schreibersite (0.32 per cent.), besides calcium chloride (0.42) and oldhamite (0·18) are deducted, the following figures given under 3a, or, calculated per hundred, under 3b, remain for the silicates soluble in hydrochloric acid.

These figures do not allow of a calculation as an orthosilicate; even if one wishes to assume that the percentage of ferrous oxide determined by indirect calculation has turned out far too low, the percentage of silica still remains so high that it can scarcely be explained by this cause alone. A determination of all the silica made as a check resulted in even several tenths per cent. more (38-71). I am sorry that I must postpone the attempt to find an explanation. But if, as from the microscopic examination I must assume, an olivine-like substance is really present, in spite of the large excess of silica, it follows, on the one hand, that this substance diminishes in favour of the enstatite, and, on the other hand, that fayalite and monticellite silicate are strongly represented in it.

4. Carbon determination.

The proportion of carbon was ascertained as 0.36 and 0.38 per

* In this I have not as yet succeeded. For if the 5.85 FeO, 1.45 MgO, 1.33 CaO, 0.33 MnO are reckoned as constituents of olivine, they will require respectively 2.44, 1.09, 0.71, 0·14 of SiO2, making a total of 4:38 of SiO2. Then there would still remain 5·40 SiO2, which could not be taken as quartz since it is the soluble part which is in question. For calculating a felspar the CaO and SiO, would suffice, but not 0·07 Al2O3. The analysis is not therefore entirely explained, unless the presence of another mineral rich in silica be assumed: but this the microscope did not show.-C. K.

cent.

Two other carbon determinations, made with black chondritic meteorites for the sake of comparison, gave, in the case of the MacKinney meteorite, 0.13, and in the case of the Farmington, 0.12 per cent. of carbon. These two figures are so low that one can draw no certain conclusion from them of the presence of a carbon or a carbon compound, while this should be the case with the St. Mark's stone.

5. The solution obtained by boiling with water contained, besides a small quantity of silica (0-07 per cent.)

One can therefore assume the presence of some calcium chloride (0-42 per cent.) and calcium sulphide (oldhamite) (0·18 per cent.); the latter was shown by Borgström to occur also in the Hvittis chondritic meteorite. From the data thus obtained there results the chemical composition given below under 6, and the mineralogical composition given under 7; in this, however, the nature of the silicates soluble in hydrochloric acid must, for the time being, be left undecided.

*

As the result of the complete investigation, St. Mark's is a black, carbonaceous quartz and plagioclase bearing enstatitic chondritic meteorite.-C. K.

Without regard to possible contained alkalies.

DESCRIPTION OF THE PLATES.

PLATE I.

FIG 1.-Shows the roughly trapezoidal, and slab-shaped form. The drifteffect developed over the whole principal surface shows clearly only on the more strongly illumined parts. In front, below, is the slightly inclined surface over which the drift-effect is continued uninterruptedly wherever the crust is preserved, so that these two surfaces together form the front of the meteorite. The bright edge on the left has been chipped and covered with a poorly developed secondary crust. The other surface is inclined at almost right angles to the principal surface, and one can see the sharp edges in which these meet. FIG. 2.-The figured side, almost at right angles to the principal surfaces, has a finely wrinkled crust with no drift-effect. Below, on the right, is seen the place where the wedge-shaped piece has broken off; here, on the two bevelled edges on the right and left, as well as on the two arms stretching over the surface, the stone shows a secondary crust.

PLATE II.

FIG. 1.—The back, uniformly covered with scar-like impressions and wrinkled primary crust; towards the edge traces of drift-effect are to be seen. On the lower edge the secondary crust can be clearly distinguished from the primary

crust.

FIG. 2.--Front, with shallow scar-like impressions and drift-effect radiating from the crown uniformly in all directions; on the right half the slight depression of the surface is observable. The darker parts at the edge are covered with primary, the lighter ones with secondary, crust. The drift-effect of the primary crust on the slightly sloping side on the right is in reality much sharper than the photograph shows.

PLATE III.

FIG. 1. Shows the structure, slightly enlarged.

FIG. 2.—Enstatite chondrule with the confused radiated structure; an example of the chondrules most common in this stone.

FIG. 3.-Vertical section through the crust. The three zones are indicated only in the middle portion bounded by two cracks, and not clearly there. Still, one can see that silicates are sparsely present, and are of smaller dimensions than in the matrix.

FIG. 4. Two nickel-iron chondrules. Of these, one encloses a small enstatite chondrule, the other isolated enstatites. The chondrule next to the latter is one of the rare ones, which are made up of a few broad enstatite crystals.

FIG. 5.-Chondrule, with diverging radiated structure, with indentations and numerous inclusions of iron pyrites. The section was treated with copper sulphate solution, and the surrounding nickel iron appears, in consequence, in larger connected particles than is actually the case.