In an attempt to obtain information about garnet nucleation a series of eleven apparently homogeneous metamorphic rocks, from a small area near Mallaig, Inverness, Scotland, have been analysed to determine bulk chemical and modal compositions, garnet grain size and number of garnet crystals occurring in unit volume of rock (N). No correlation of N with any function of chemical and modal compositions could be found, and it is therefore concluded that, if the influence of trace components can be neglected, externally imposed factors (for example, time-temperature variation) or internal physical factors (for example, the surface area of the rock just prior to garnet formation) must exert a controlling influence on garnet nucleation. It was found, however, that in each specimen, there was a preferred crystal radius and the relative frequency of occurrence about this preferred value was constant for all samples studied. Preliminary comparison with other systems, investigated by the authors, suggested that the nature of this relationship may provide evidence of the conditions under which these rocks recrystallized.

Certain amygdaloidal tholeiitic rocks and associated volcanic breccias are intrusive into the Cuillin gabbro (sensu lato) and are not xenolithic as Harker believed. Six tholeiites have been analysed and their compositions compared with the non-porphyritic central magma type of Mull. It is postulated that the ultrabasic rocks close to some tholeiitic intrusions were contaminated by the tholeiitic magma.

Specimens of Parabrograptus and Sinograptus from western Canada are described for the first time. An amended diagnosis of Parabrograptus is given. One pyritized fragment of a sinograptid stipe exhibits characters intermediate between Holmograptus and Sinograptus.

The age of the Garabal Hill-Glen Fyne igneous complex has been determined as 392 ± 4 m.y., by the Rb-Sr method. The components of the complex had an average initial (87Sr/86Sr)0 isotope ratio of 0·705 ± 0·003. Geological contamination of marginal components is reflected in the (87Sr/86Sr)0 values of these rocks which were found to be ca. 0·710, significantly higher than values from uncontaminated samples. The results are in accord with the view that the individual members of the complex had a common source, comparable with regard to the Rb/Sr ratio, to the source of basalts. It is inferred that complete melting of such source material followed by magmatic differentiation is an appropriate mechanism of origin for the complex.

Three large pipes of brecciated Torridonian sediments set in matrices of calcite and calcareous mudstone occur on the island of Raasay. Previous work has shown that these structures may be due to explosive fumarole activity during the Tertiary volcanic episode. A recent study of the Torridonian stratigraphic succession has made it possible to demonstrate that there was considerable variation in the amount of vertical collapse of the local country rock caused by the explosions. In one vent local fracturing and brecciation took place with no downward movement; in another it is possible to show that Torridonian boulders tumbled at least 350 metres down into the throat of the fumarole. The amount of vertical collapse increases from vent to vent southwards towards the Cuillins volcanic centre.

These vents are of particular interest for two reasons. Though clearly of volcanic origin they contain no volcanic material. Secondly, the carbonate matrix of analogous structures is generally believed to be due to the hydrothermal solution of underlying limestone beds. The latter are absent in these examples and the carbonate material seems to be solely of volcanic origin.

Early folds (F1) in the Mylor Beds are upright or overturned to the north-west and trend north-easterly. They are cut by a flat-lying cleavage (S2) which is related to small-scale recumbent folds (F2) superimposed on the steep limbs of the early folds. The sequence of development of folds can be found elsewhere in the Mylor-Gram-scatho Beds and also in the Lizard Head mica schists. It is shown that F2 folds and S2 cleavage resulted from vertical shortening, which is believed to have been caused by the superimposition of slide sheets. It is suggested that the previously accepted age order of the Mylor and Gramscatho Beds should be reversed.

Arising from detailed reassessment of field work in the area of Ny Friesland and Olav V Land from 1938 to 1965, a revised stratigraphical scheme is proposed for the Hecla Hoek succession which is known to range downward from Lower Ordovician and Lower Cambrian (1 km.) into a relatively unbroken sequence of Pre-Cambrian rocks about 19 km. thick. The metamorphosed older rocks, previously grouped as Lower Hecla Hoek, are reinterpreted as a succession 12 km. thick with a large volcanic component and a distinctive tillite horizon. Correlation is attempted anew with rocks elsewhere in Spitsbergen and overseas. The Hecla Hoek suite contributes to an interpretation of the development of the Caledonian tectonic province in late Pre-Cambrian and Lower Palaeozoic time.

Distinct variations of the natural remanent magnetism in a section of volcanic rocks in Greaser Canyon, Lake County, Oregon, correlate with faulting inferred independently. The palaeomagnetic data is also consistent with the occurrence of a suspected diastem in the local volcanic sequence.

The model recently proposed by Kemežys for the distribution of punctae on the brachiopod shell is examined and found to be insufficient. An alternative model is proposed: punctae appear to conform to a hexagonal close-packing pattern on the shell surface. A hypothesis to account for this is introduced, and further implications are discussed.

Joints (fractures) are propagated from areas of high to areas of low stress. They may originate at inhomogeneities on, or by slickensiding along bedding planes. Plumose marks on the surfaces of joints radiating from volcanic, meteor, or man-made explosion craters should, in stratified rock, reveal whether the joints were propagated with their leading edge convex or concave forward.