The main thesis of this paper is to show that the Ludlow Bone-bed forms the true base of the Devonian System. This is urged on several grounds.

(a) Priority. It was the limit originally used by Murchison. Although subsequently altered by him, one finds if one attempts to use his later limit:—

(i) That the Old Red Sandstone was not re-defined accordingly.

(ii) That Dumont's Rhénan (Devonian) System, defined in 1845, has priority over Murchison's later definition (1848).

(iii) That the later limit adopted by Murchison does not form a true horizon and has been very differently interpreted by later writers.

(b) Palæontology.

(i) There is a marked faunal change at the Ludlow Bone-bed. The Bone-bed marks the first appearance of fossil fishes, and thus the Devonian marks the dawn of an age of vertebrates. There is only one old record of a pre-Devonian fish in the British Isles; one doubtful one in Gotland, and an anomalous one in the Ordovician of America. The Shropshire succession forms a standard of comparison for all areas of the Old Red Sandstone type of deposit.

(ii) The limit adopted after a detailed study by other writers of an uninterrupted marine series is on the same horizon. The succession at Liévin (Northern France) forms a standard for comparison of all areas of the Devonian (marine) type of deposit.

(c) Stratigraphical. In England the Ludlow Bone-bed marks also a change in physical conditions, more noticeable elsewhere as an unconformity. Physical breaks or unconformities on the same horizon occur in Scotland, Belgium, Brittany (slight), Norway, Sweden, and Spitsbergen.

(d) Practical Considerations. The Downtonian rocks, i.e. the strata from the Ludlow Bone-bed horizon upwards, form a stratigraphical and palæontological entity, incapable of broad or even of rough separation below the Dittonian. The Ludlow Bone-bed, on the other hand, forms an horizon which can be recognized and mapped all over North-Western Europe. In England there appears to be a slight faunal break between the Downtonian and Dittonian, but this is less marked elsewhere.

Rhombic pyroxene as a constituent of thermally altered sediments has hitherto been recorded only from a few contact zones, and, as the conditions under which it is generated in metamorphosed sediments are of some interest, these may be examined.

In an earlier paper (8)an olivine-basalt was described from this locality; the present contribution may be looked upon as a continuation of the study then begun. This basalt is best exposed at the southern end of the ridge in a small quarry, but several exposures are to be seen farther north. All these, where the compact, fresh-looking basalt is to be observed, are situated along or near the eastern margin of the ridge, a fact particularly well shown near the northern end. On the western side of this strip are several exposures of fine-grained, brownish rock, so well jointed that it is almost impossible to get a good specimen; an additional feature of importance is that the exposed surfaces are frequently striated. That these striations are not of glacial origin is demonstrated by their occurrence on newly-quarried surfaces, and also by the fact that specimens bounded on two sides by these striated surfaces may be collected.