The rock surface to be described in this paper was discovered under boulder-clay during excavations for a new reservoir for the City of Cardiff Waterworks at Rhiwbina, 5 miles north of Cardiff. The relation of rock to boulder-clay proved so unusual, by reason of a curious jointing in the rock-surface, that it appeared to be worthy of record.

The main conclusions of the paper may be briefly stated thus:—

1. The deposits, which are of neritic facies, possess a characteristic and peculiar mechanical and mineralogical composition, differing therein from any other British deposit.

2. They exhibit a general constancy in grade from the Dorset coast to the Cotteswolds. Within this limit they conform to two chief types containing respectively 40 to 50 and 60 to 80 per cent, of the very fine sand grade (greater than 0·05 and less than 0·1 mm.). No variation of mechanical composition with hemeral change can be observed.

3. The sands are blue and glauconitic in depth. At the surface they are decalcified and are yellow and brown as a result of oxidation. Minerals such as pyrite and pyrrhotite occur only in depth.

4. The mineralogical composition indicates an abundance of highly angular brownish-pink to colourless garnets (possibly derived from rocks like those of the Lizard), but few red or purplish-red varieties; also of micas (including a pseudo-uniaxial, pale-green to colourless variety), together with chlorite, chloritoid, kyanite, staurolite, orthoclase, and microcline.

5. Tourmaline is only moderately common and is always the grey-brown variety.

6. The abundance of titanium-minerals is characteristic of the deposits. Red and yellow varieties of rutile and sagenite-webs are exceedingly abundant, and anatase, brookite, and sphene are locally so. A reciprocal relation may exist between the occurrence of sphene, ilmenite, and the oxides of titanium.

7. Epidote and glaucophane are rare.

8. Chlorite, chloritoid, glaucophane, kyanite, and staurolite increase in quantity as the deposits are traced southwards.

9. A change in lithology and mineralogy sets in as the deposits are followed north-eastwards into the Midlands and Yorkshire.

10. The sands differ markedly in petrology from the various Palæozoic rocks of Wales and the West Country, from the Trias, and from the Cretaceous and Eocene of Devon and Dorset.

11. The Ordovician, Domerian, Toarcian, and Aalenian of Normandy and Brittany have been examined petrologically for purposes of comparison, but show little resemblance to the sediments under discussion, having been derived locally from the Palæozoic and pre-Cambrian rocks.

12. It is concluded that the Lias–Inferior Oolite sands have been derived from the south or south-west, notably from rocks like those of western Brittany.

13. The rise of the Mendip axis was not sufficiently complete at this time to cut off the supply of sediment from the south or south-west. The deposits north and south of the ridge have the same general character.

I have for several years paid attention to glacial phenomena near Coniston, and some of my observations seem to be worth recording.

The majority of the crinoids recorded in this paper are from Penton Linns and a few from Harelawhill and Peterscrook in Liddesdale on the Scottish border. Peterscrook is really on the English side of the Liddel, which here forms the border line, but being quite a short distance from the other localities, it is thought advisable to include it here. The specimens in my own collection are the result of five visits at various times during recent years, the first of which, in the autumn of 1920, I was accompanied by my lamented friend, the late Mr. Robert Dunlop. My companion on the three following visits was Mr. James L. Begg, of Glasgow, and on the fifth visit Mr. J. G. Duncan, late of the Royal Scottish Museum and now of Kirkcaldy. To our combined efforts the following results are due. An account of these results may be of interest more especially since the crinoid fauna bears a remarkable resemblance to the “Woodocrinus” fauna discovered many years ago at Richmond in Yorkshire and first made known to geologists in the pages of the Geologist and later in the Geological Magazine.