In the Harlech Grit Series pyrite always occurs associated with rocks of practically all grades. It is, on the whole, of least importance in the grits, but even in these it becomes locally fairly abundant. Among the mudstones it occurs concentrated into layers, which are relatively persistent laterally. Immediately below the manganese-ore of the Manganese Shale Group, for example, there occurs a very widespread band of green mudstone which contains abundant pyrite crystals. Although its general characters are the same this band varies slightly, in detail when traced laterally and it is with these variations that the present paper is largely concerned. This band is usually 1 to 2 inches thick and is usually of dark green mudstone, though rarely it is of the normal blue-grey colour. The pyrite crystals vary from very minute to about ⅜ inch in size. In addition, pyrite occurrences from other rocks are described. The account is particularly concerned, not with the pyrite crystals themselves, but with zones of authigenous minerals which almost invariably surround them in these rocks.

It has long been suspected that the mode of development of the proximal end in Cryptograptus may differ considerably from the ordinary Diplograptid type, but the preservation of the material has hitherto been too poor to allow of a complete interpretation. Amongst the graptolites preserved in impure limestone nodules1 in the Balclatchie Shales at Laggan Burn, near Girvan, Cryptograptus tricornis is abundant, and despite the extreme tenuity of the periderm, specimens can be dissolved out with HC1 and HF, in a remarkably good state of preservation.

AMMONITES VARIANS J. Sowerby2 is a composite species, based on five syntypes, which are preserved in the British Museum (Natural History),3 and have long been accessible to any reviser who wished to select a lectotype from among these syntypes. Nobody, however, has yet done so. In the days of the early revisers, of course, from Mantell and Brongniart (1822) to Sharpe (1853), nomenclatorial restrictions were unknown; and even when Semenow (1899)4 dealt with Sowerby's species, he accepted the customary interpretation of Schloenbachia varians without going back to Sowerby's original description and figures. Semenow took d'Orbigny's5 excellent figure as typical, and his restriction seemed to me convenient to adopt,6 if not actually binding. But on going more fully into the matter I discovered not only that Semenow's selection (if it be selection) is not valid but also that none of the forms which authors nowadays identify with Schloenbachia varians agrees either with Sowerby's description or with any of the syntypes in the British Museum. It is thus highly desirable to discuss this common zonal ammonite species and place its interpretation on a sound footing.

This paper presents a brief outline of the geology of the area shown in the map, Text-fig. 1, and a petrographic account of a small collection (eighty specimens) of its igneous rocks. The area is centred about 60 miles south-south-east of the southern end of Lake Tanganyika, and midway between Abercorn and Kasama. It lies at the headwaters of the Chambesi River, the source, or one source, of the Congo, on the high plateau of Northern Rhodesia at an altitude of 4,000 to over 5,500 feet. There is as yet no literature dealing with the petrology of North-Eastern Rhodesia, and the Rhodesian Copper Belt is too far distant (over 300 miles) to allow of any useful comparisons at this stage. It is interesting, however, to compare the rocks of this area with those of the Lupa Goldfield (1) in Tanganyika Territory, some 120 miles to the north-north-east.