The fragments of jaws of Ornithosaurs from the Cambridge Greensand show a greater variety in size, form, and proportion than those from any other formation. The largest species were apparently the most singular in the shape of the snout, but after five-and-twenty years of collecting we still seem destined to know them only from fragments which, though extraordinary and suggestive, are tantalizing from their imperfect condition. It is impossible to tell whether the head was short or long, or what proportion it bore to the body, in the oase of these isolated specimens, but some of them, like the jaw-fragment which I am about to describe, show a power of tooth and massiveness of bone which could only have pertained to one of the most destructive and probably one of the largest of these animals.

It was indicated in a previous paper (Geol. Mag., LXXII, 1935, p. 208) that the Scottish Lower Carboniferous crinoid fauna contains a number of species which have never been adequately described. In the present communication, ten of these are discussed, the remainder, another ten or eleven species, will be dealt with later. All the specimens except four were collected by myself many years ago, but it was only recently that the requisite leisure was found to develop and clean them thoroughly.

The Kraemer Island macrodyke that is exclusively exposed on Kraemer Island about 7 km west of the Skaergaard Intrusion belongs to a regional dyke swarm termed the ‘Skaergaard-like dykes’ (or FG-1 dykes). Weakly modally layered olivine gabbros dominate the exposed parts of the intrusion that has a width of 650 m to 1000 m. Plagioclase (core An68±2) and Ca-rich pyroxene (core Mg no. 79±1) grains are normally zoned, whereas olivine grains (Fo50–65) are homogeneous. Calculated mineral–magma equilibria, based on experimentally determined Mg–Fe magma–olivine and magma–clinopyroxene partition coefficients, suggest that the observed olivine and clinopyroxene compositions in the gabbros cannot have formed from a common parental magma. The unzoned nature of olivine grains and their iron-rich com-positions relative to clinopyroxene suggest post-cumulus Mg–Fe exchange between olivine and interstitial melt. A gabbroic pegmatite is developed in the centre of the intrusion along its entire exposed 5 km strike length. Here, mineral zonation is limited and compositions are similar to rims of cumulus minerals in the enveloping olivine gabbros. The pegmatite could therefore represent interstitial melt mobilized from gabbroic cumulates that later accumulated and crystallized at its present stratigraphic location. Cumulus olivines in the gabbros are close to equilibrium with Ca-rich pyroxene in the pegmatite. This is interpreted as reflecting interstitial melt mobility during the late stages of solidification of the macrodyke. Chilled margins are well preserved at the intrusion margins and are rich in Fe (14–15.3% FeOTOT), Ti (3–3.3% TiO2) and light rare-earth elements ([La/Sm]N = 1.2–1.3), similar to magmas well represented in the overlying sequence of contemporaneous High-Ti Series flood basalts. It is therefore likely that extensive wall-rock melting adjacent to the macrodyke reflects continuous feeding of the overlying flood basalts through the Kraemer Island macrodyke.

An occurrence of kyanite pseudomorphing andalusite, from the Urungwe District of Southern Rhodesia, is described. The pseudomorphs, which include some of the largest of their kind so far recorded, occur in profusion in certain Pre-Cambrian pelitic schists. The original andalusite was completely converted to kyanite during regional metamorphism but the shape of the original crystals was perfectly preserved. This, together with the habit of the kyanite within the pseudomorphs, shows that the change took place under conditions of insignificant shearing stress.

Sedimentation ages of metamorphosed clastic sedimentary rocks in the southern Schwarzwald were determined by associations of palynomorphs. In the northern subunit of the Badenweiler–Lenzkirch Zone, two lithostratigraphic assemblages could be discerned in low-grade metamorphic units by their facies and age, thus revealing a more complex internal structure of this zone than previously assumed. Lower Ordovician metagreywackes and metapelites were discerned from Silurian metasiltstones. In the cataclastically overprinted metasiltstones and phyllites of the southern subunit of the Badenweiler–Lenzkirch Zone, only poorly preserved microfossil remains could be detected. These show that the sedimentation ages must be Ordovician or younger, but still probably Early Palaeozoic. High-grade metapelitic rocks of the South Schwarzwald Gneiss Complex contain chitinozoans in lenses and layers of schists, that are rich in biotite and graphite. They yielded mid-Silurian ages and show that this crystalline complex does not represent an older basement unit but was the result of marine sedimentation at that time. The new age determinations have a bearing on geodynamic reconstructions of the internal Variscides in Early Palaeozoic time. They show that sedimentation in the oceanic realm of the Badenweiler–Lenzkirch Zone or its margins did not occur before the Ordovician. After transformation of the northern passive into an active continental margin, younger greywackes not older than Middle Devonian received detritus from a volcanic arc, forming above the subduction zone.

Micropholis Stowi was described by Huxley in 1859 from a small and very incompletely preserved skull found by G. W. Stow at Rhenosterberg (north-west of New Bethesda), District Graaf Reinet, Cape Colony. Subsequently R. Owen described another specimen as Petrophryne granulata. In his description he suggested that it might prove to be identical with Huxley's type. The British Museum now contains these two type-specimens and three other examples of the form, all except Huxley's type being from the Procolophon zone of Donnybrook, Upper Zwartkei, District Queenstown.

A number of test bores were sunk for Swansea Corporation in 1931 during the investigation of a site. Some samples were obtained by Professor O. T. Jones, F.R.S., from a line of borings east of Swansea, which trended north-east from the eastern end of King's Dock and Queen's Dock.

Parallel folds exhibit a characteristic orthogonal relationship between the tangent and the corresponding isogon drawn at any point on folded surface. Modification of parallel fold to flattened parallel fold by superimposition of homogeneous strain introduces an angular shear along the tangents at different points. The angular shears in different directions, obtained by measuring angles between the tangents and the corresponding isogons, can be used for estimation of flattening strain by a variety of geometrical and numerical methods. We show that several simple geometrical techniques, such as the Wellman method and the Mohr circle method, can rapidly decipher the strain from flattened parallel folds. These methods, in contrast to most of the existing methods of strain estimation, are independent of the assumption that one of the principal strain directions parallels the axial trace on the profile plane of fold.

The progress of modern discovery, both in Palæontology and Zoology, has been steadily tending to bridge over the gap which seemed at one time to separate the past from the present life-history of our globe, and has brought each more closely into relationship with the other than seemed possible fifty years ago.

Mudstone inclusions in graded greywacke-conglomerates of Salopian age from the Slieve Bernagh Mountains, Co. Clare, are considered to have been formed by fracture of the conglomerates in the course of slumping shortly after deposition and the infilling of the resulting cavities by mud derived from the underlying beds. The attitudes of the inclusions and of associated pebble imbrication suggests that slumping took place on a slope with a northerly inclination of 10°.

An interesting occurrence of highly potassic lavas and hypabyssal intrusives of assumed Tertiary age is described from North Johore. The thickest lava flow is gravitatively differentiated showing a basal horizon of potassic ankaramite and basanite grading up into highly potassic bolitized flow tops of leucite-tephrite composition. Evidently the parent magma was close to a biotite pyroxenite in composition and possibly an assimilation product of peridotite by granite.

Although the ammonoid Durvilleoceras is apparently very close in its morphology to early Triassic genera, it comes from a formation that underlies formations with early Triassic and late Permian faunas, and appears to be of late Middle Permian age. New occurrences of the ammonoid support this thesis. Conjecturally, the genus may have inhabited deep cold waters of the southern hemisphere during the Permian Period, before giving rise to genera found in shelf deposits of the early Triassic. Alternatively, if really Triassic in age, Durvilleoceras indicates a major low-angle thrust, previously unsuspected, that has repeated Triassic sequences for a length of over 450 km before disruption by the Alpine Fault. No evidence is yet known to support this alternative.

Having described in some detail, the rocks of the Maentwrog, Festiniog, and Dolgelly groups, and given a list of the fossils found in each, I shall now make some general remarks upon the fauna. In Part I. I have mentioned that out of all the numerous Trilobites from the Lower Cambrian rocks, only two genera, Agnostus and Conocoryphe, have been found in higher strata. Strictly speaking, this is only true of Agnostus, which passes unchanged in type from the Menevian to the Caradoc strata. On the contrary, the species of Conocoryphe from the Lower Cambrian rocks are of essentially different type from those that have been placed in that genus from higher beds. The latter have, I believe, been referred to the genus Conocoryphe rather from superficial resemblances than from true affinity.

Allogenic biotite flakes, together with fish and conodont remains, are concentrated into bands at particular horizons within Upper Ludlovian rocks of Woolhope and other Welsh Borderland areas; and are present but more scattered within the Aymestry Limestone and Dayia Beds. Metamorphic rocks of Malvernian type appear to have provided the source of the biotite.

More than twenty years ago, whilst engaged in the study of the Lower Carboniferous rocks of Westmorland, I noticed the occurrence of certain small concretionary nodules of very compact texture in the dolomites near the base of the succession in the neighbourhood of Shap.

The incommodus, atavus and cyphus monograptid groups appear more or less simultaneously in the atavus Zone. The atavus and cyphus groups are related, one of these probably giving rise to the later tenuis group of monograptids, Possible roots of these groups and relationships with later monograptids are discussed. Monograptus strachani sp. nov. is proposed to replace the name M. incommodus sensu Elles & Wood.

The material upon which the following descriptions are based was obtained chiefly by Mr. Ussher, who collected the Whiteway and Hestow Farm specimens. The Daymer Bay form was collected by Mr. John Pringle. For the new species of Proetus I am indebted to Mr. Claude H. Peter, Town Clerk, Launceston, who, through Mr. Clement Reid, kindly lent it for description. It is evident from the examination of this limited quantity of material that the British Devonian fauna is not yet fully known. We may expect the discovery of many species hitherto unrecorded in this country, which will be of interest from the possible comparative relationship with the fauna of foreign deposits. The accompanying illustrations were drawn by my colleague, Mr. H. H. Thomas, to whom I tender my best thanks.

The specimens here described were obtained by the writer from a pocket in the surface of the Silurian Slates about 200 yards to the north of Gillet Brae, exposed during the preliminary excavation for Mr. Delany's quarry (the Horton Lime Works) in 1889. The quarry is situated on the right bank of the Ribble, opposite Horton village. The site of the exposure is now buried under tip heaps and there is little chance of further sections being exposed.

Although numerous papers, principally by the late E. Wilson, describing the beds of the Rhætic and the Lower Lias of Notts, were published while exposures existed along railway cuttings in the south-east of the county, no account of the precise positions of important fossils is to be found. These exposures are now all grass-grown, but the junction of Rhætic and Lias may be seen in a number of quarries where hydraulic limestone is worked, either for cement or for road material. One of the best is at Owthorpe. Those at Cotgrave Gorse, Barnstone, Plumtree Wolds, and Normanton Hills are confirmatory. The layers rich in Foraminifera have been of value in correlating these sections.

—The field impression of a high degree of uniformity in the composition of the intrusive Karroo dolerites over a vast area is confirmed by new chemical and microscopical studies. The pyroxene characteristic of the dolerites is a typical pigeonite, which, however, does not display any twinning on (100). The plagioclase is potash-free, usually a labradorite, but it has been observed that a homogeneous crystal of plagioclase may be in twin position to another homogeneous crystal of a very differently composed plagioclase. The close resemblance of the doleritic liquid to the artificial liquid taken by Bowen to represent basaltic liquid is noted, and his conclusion that pyroxene and feldspar should crystallize simultaneously in natural basaltic liquid is confirmed.

Conceivably the doleritic liquid was quite original and not a differentiate of any earlier liquid. However, analogies like those with the tholeiites and similar hypabyssal rocks of Great Britain suggest that in South Africa, as in Great Britain, the liquids of these hypabyssal rocks were derived from the slightly more femic plateau-basalt.

On that assumption the question of the mode of differentiation arises. The fractional crystallization of plateau-basalt, as now described by Bowen and other leaders, does not appear competent to explain the abnormally low soda of one of the analyzed dolerites, nor the excess of soda and total alkalies in plateau-basalt respectively over the soda and total alkalies of the Karroo dolerite. Further, the settling-out of early olivine does not explain the excess of (total) FeO in plateau-basalt over the (total) FeO in the Karroo dolerite. The actual relations indicate the need of renewed examination of the theory of magmatic differentiation by crystal-fractionation. In any case many more data are required before it is possible to decide upon the precise relation of the Karroo dolerite to a parental magma. Possibly such additions to knowledge may annul present difficulties in the way of accounting for the composition of the dolerite.