The extinct volcano of Katunga (E. 30° 11′ 27″; S. 0° 28′ 17″ approx.) is in northern Igara which is part of western Ankole in south-west Uganda. The volcano is of special interest: (a)because of its isolation, the nearest neighbouring manifestation of volcanic activity being that in connection with the crater in which Lake Nkuguti lies, about 12½ miles to the N.N.W. in Bunyaruguru; and (b) because true lava-flows, the first to be discovered in the volcanic areas of north-west Ankole and Toro, composed of potash-rich olivine-melilitite (katungite), were extruded from it. The volcano is an outlier of the southern edge of that part of the western volcanic field of Uganda lying in, and adjacent to, the portion of the western rift depression about the northern end of Lake Edward, the southern end of the Buwenzori massif, and Lake George.

On the eastern side of Barnavave, at the immediate junction of the igneous mass with the surrounding limestone and about a quarter of a mile south of the fault which crosses the mountain, there is an isolated outcrop of a fine-grained, speckled rock which clearly invades the limestone at this point. The rock includes xenoliths broken from the limestone and also others of igneous origin.-and smaller dimensions. Microscopical examination proves the rock to be a syenite of unusual character which carries a small amount of quartz. It is possibly the rock referred to by Harker.

The Composition of the Sand.

The Lower Bunter consists of sand whose average grain size is about ·2 mm. Particles below 0·05 mm. or above 1 mm. are virtually absent.

The fineness of the sand and restricted limits of size variation in the whole thickness of the deposit, and the high perfection of sorting in individual samples indicate an aeolian origin.

The False Bedding.

No true bedding exists in the Lower Bunter examined.

The false-bedding is only explicable on an aeolian origin. It is argued that it must represent cross-sections of barchan dunes advancing in a sand sea that was steadily increasing in thickness.

The frequency distribution of the values of the false-bedding dips confirms this interpretation.

The directions of different types of false-bedding indicate that transport and deposition was effected almost entirely by a prevailing east wind.

Relation of the Lower Bunter to the Mercian Highlands.

The establishment of an easterly origin for the sand indicates that the “ Mercian Highlands ” of the Birmingham district were the source of sand supply.

Arguments are adduced towards the correlation of the Barr Beacon Beds and the Lower Bunter Sandstone, the former being regarded as the thin piedmont spread of scree material fringing the desert.

A Number of small pegmatitic patches occur in the development of the later stages of the hornblende hybrids of the Cairnsmore of Carsphairn Complex. These are best seen on the eastern side of the Gairy of Cairnsmore, and it is within these pockets of late residuum crystallization that this hornblende occurs. The pegmatites consist of milky quartz, greenish-black hornblende and a clove coloured sphene. The two latter minerals develop idiomorphic crystals. The hornblende, though usually only 1½ inch in length, occasionally grows to 1½ inches long in slender prismatic crystals showing the faces (100), (110), (110), (130), (130), (010), (011), and (011).

The name of this species was proposed by Schlotheim (1832, 65, No. 88), but the history of the form really goes back to Linnaeus. The earlier writers on the Scandinavian Palaeozoic took the name Anomia plicatella Linn. (1758, 702) to cover a great number of different ribbed brachiopods, and referred in their synonymies to the form figured in fig. 5, of tab. v, of Linnaeus's Museum Tessinianum (1753), a work published some years before the adoption of binomial nomenclature. When different authors began to subdivide this group, the name plicatella came to be restricted to a species referable to the genus Spirifer, and though some authors, Wahlenberg (1821, 67) and Dalman (1828, 140, Tab. vi, fig. 2), continued to use the name plicatella in a wide sense, so as to include the species under consideration, the latter seems to have come within the range of forms described by Schlotheim (1813, 36) as Terebratulithes lacunosus. This author apparently had the intention of separating a group of ribbed Silurian brachiopods, under the name of Terebratula borealis, for in the catalogue of his collection (Schlotheim, 1832, 65, No. 88) six specimens from the “Uebergangskalkstein” of Sweden are entered under this title. The term, however, remained a nomen nudum until von Buch (1834) gave an adequate description, and it is to this author that the species should correctly be ascribed. He chose as type a specimen from the Encriniten-Übergangskalkstein of Gotland, figured in an earlier work by Schlotheim (1822, 68, Taf. xx, fig. 6a—c) as Anomia Terebrat. lacunosus.

The specimens representing the Katunga lavas are compact fine-grained porphyritic rocks of lustrous stone-grey to deep blue-grey colour, the phenocrysts being olivine (up to 3 mm. long) of various tints from nearly colourless to yellow-green, and melilite (up to 2 mm. square) in thin tabular, less conspicuous crystals which, megascopically, have a shining grey to black appearance. Some of the specimens from near the sources of the flows are internally altered and have become fawn or buff in colour, mottledby white zeolitic specks and reddish brown ferruginous pseudomorphs after olivine and melilite. Some weathered specimens have a brown ochreous crust, but within this the rock is normally fresh.

The specific name of this form is not in doubt, but its authorship is less securely determined. The species Terebratula cuneata was proposed by both Dalman (1828, 141, Tab. vi, figs. 3a–c)and Hisinger (1828, 220, 239, Tab. vi, fig. 5) in the same year, and the first time each author mentions the name, he ascribes it to the other. Each account is accompanied by figures, and there appears to be no ruling of the International Commission on Nomenclature that would settle the authorship in such a case. Dalman's account, however, appears in the volume for 1827, though the title page bears the imprint 1828, and such a choice of authorship agrees with alphabetical priority and with the usage of most subsequent palaeontologists.

The deposit to be described is situated at Orleigh Court, in the parish of Buckland Brewer, some four miles west of south from Bideford, where it rests unconformably on the steeply dipping Upper Culm sandstones. Its greatest extent is from Orleigh Mill to Yeo Bridge, a distance of about three-quarters of a mile in a northwest to south-east direction, and from this line it extends to the south-west for an average distance of about one-third of a mile, only exceeding that amount to the south of the Higher Lodge. The only extensive section was seen in the Rookery, where 25 feet of material rest on a yellow clay. The highest point reached by the deposit, to the south-west of the Higher Lodge, is about 400 feet above O.D. From this point it slopes to the north-east, extending down to about the 100 ft. contour on the left bank of the River Yeo. It is possible that the north-eastern boundary does not reach such low levels as those indicated on the map, Text-fig. 1, for it is difficult to distinguish the original gravel from relatively recent hill wash; and, without a good deal of trenching or augering, it would not be possible to determine the boundary more closely than has been done in the present work.

(1) The Elsworth Rock of Elsworth and St. Ives is a condensed deposit belonging to the plicatilis zone.

(2) There is no evidence that it contains any ammonites derived from earlier zones. Statements that it does are based on misidentifications.

(3) The Elsworth Rock at Elsworth overlies at least 8½ feet of ironshot clays with thin bands and nodules of limestone, the clays containing abundant ammonites of the cordatus zone and no others.

(4) The so-called Elsworth Rock at Upware is the equivalent of the ironshot cordatus zone developed below the Elsworth Rock at Elsworth.

(5) The true Elsworth Rock (plicatilis zone) is represented at Upware by the Coralline Oolite and Lower Coral Rag.

(6) For convenience in stratigraphical nomenclature it is desirable to group together the condensed ironshot cordatus and plicatilis beds of Elsworth as the Elsworth Series.

(7) The Elsworth Series, as defined in 6, corresponds with the Oolithe ferrugineuse and the greater part of the Oolithe de Trouville on the Normandy coast. At Trouville the plicatilis zone to some extent combines the developments at Elsworth and Upware.

(8) The ironshot cordatus zone was proved in the Upware well to rest directly upon normal clays of the mariae zone, and the same zone was proved in the vale below Elsworth in the Fen Stanton brewery boring.

I am greatly indebted to Mr. E. E. L. Dixon and Mr. A. G. Brighton for criticisms and suggestions, and especially to Mr. Brighton for the trouble he has taken in sending me all the fossils on loan and supplying the relevant information.

The marginal rocks of this intrusion were described briefly in a paper dealing with the Cairnsmore complex. The intrusion is composite, consisting of a central granite, and successive zones of acid hybrid, tonalite, and basic hybrid. The complex began with the intrusion of gabbro, and when crystallized, but still ata high temperature, it was followed by the tonalite with the production of the basic hybrids. The original basic intrusion is nowhere found in an unaltered condition but is now represented by the “dioritic” lenses. It is seen then that the outermost zone of the intrusion uncontaminated by the incorporation of the sedimentary country rocks is itself a hybrid resulting from the intrusion of tonalite in the original gabbro. In the earlier paper it was suggested that the character of these marginal rocks was due to the combined effects of sedimentary contamination and igneous hybridization.

In a communication to the GeologicalMagazine, published in July, 1936, I gave a brief description of the rocks that build the Kedong scarp. I propose now to amplify that description. It is not always realized that the geological structure of the Rift Valley has been deduced almost wholly from physiographic evidence, and that proof of the alleged displacements must depend upon exact petrographic study of the rocks involved in the structures.

The fossiliferous rock now reported upon has been submitted to me by Mr. E. S. Willbourn, Director of the Geological Survey of the Federated Malay States, in the hope that examination of the fossils might lead to some definite conclusion as to the age of the coal-bearing strata with which it is associated. The locality is Durian Chondong, Kepong, in the Unfederated Malay State of Johore, and the material examined comprises a number of bore samples obtained from depths varying from 88 feet to 145 feet. These all consist of a grey marly shale containing numerous fossil shells, mainly more or less crushed, but sometimes well preservedand even retaining traces of their colour-ornamentation. The only other fossils obtained from the samples, beyond a few obscure plant impressions, are some small and indeterminate fish teeth.

The effect of waves on a shore depends largely on the angle of approach, for the energy of the advancing waves tends to erode the headlands and drift the products of erosion into the bays. Wave refraction applies to waves approaching the shore in any direction, but is only ideally developed in ordinary weather conditions or by approaching waves caused by distant storms. Locally developed storm waves usually approach the shore at an oblique angle to its general trend, owing to the failure of wave refraction and are the most effective where locally they meet the shore at right angles. Beach drifting will be in operation along the section of the shore which happens to be at an angle oblique to the direction of the approaching waves.

While examining the collection of Old Red Sandstone fishes in the Royal Scottish Museum, the writer found the frontoethmoid shield of a Crossopterygian fish, from Eoghole, on the Muckle Burn, near the boundary between Elginshire and the county of Nairn. Though imperfect, the skull fragment retains its natural shape, and at once recalled the similar region of Eusthenopteron, from the famous fish-bearing beds of Scaumenac Bay, Province of Quebec, Canada. Further comparison seems to prove the generic identification. The stratigraphical relationships of the find are known with some precision, and the implications are of considerable importance. It is especially opportune that Dr. Erik Jarvik, of Stockholm, has just finished an account of Eusthenopteron species from the Baltic and Russia, with a discussion of the stratigraphy of those localities which yield the fish. Certain features of the Canadian Eusthenopteron are discussed in the present paper for comparison with the Scottish form: this is in no way prejudicial to the monograph on the Scaumenac Bay material, now in preparation by Dr. Jarvik.

The well known Lulworth Cove, situated on the Dorsetshire coast between Mupe Bay, on the east, and Dungy Head, on the west, is generally considered to be a very good example of the result of marine erosion acting upon rocks of unequal resistance, and such indeed it is.

Since the publication in 1887 of Dr. H. B. Guppy's book (1) very little has appeared on the geology of the Solomon Islands, and Cowper Reed's book (2) of 1921 is based on Guppy for this area. During the last four years a small representative collection of rocks has been sent to the British Museum (Natural History) and examined by Mr. W. Campbell Smith, to whom the author is much indebted for their identification.