Few plants of Carboniferous age have appealed so strongly to the scientific observer as the huge woody trunks of a coniferous type that have been discovered from time to time in beds of that system. The appeal has come from different points of view at different times; thus these trees were hailed by Witham (1831, pp. 1–2), Lindley and Hutton (1831, p. xiii), and Hugh Miller (1849, p. 186) as evidence of the existence of higher coniferous plants in abundance during a period to which the current consensus of scientific opinion had assigned only members of the lower plant families, or would admit only an infrequent occurrence of higher forms. These remains were therefore important in discussions concerning the development of plant life, and were cited by Hugh Miller (1849, p. 185) in his attempt to combat the then embryonic ideas of evolution—the recrudescence of Lamarckian ideas—current under the name of “the development hypothesis.” Miller, however, recognised one significant point, namely, that they represented a land flora as distinct from one occupying a water habitat (1849, p. 202).

The existence in the Island of Arran of petrified Lepidodendroid trees of Lower Carboniferous Age has been known for nearly seventy years, and yet only very brief and incomplete accounts of these interesting fossils have been published. It is the aim of this paper to provide some additional facts which have a bearing on the past history of these trees, and to furnish an illustrated account of their detailed structure. It is possible from the additional knowledge that this research has placed at our disposal to get a clearer insight into some aspects of the developmental history of these Palæozoic trees.

At the instance of Dr S. L. Hora I was sent in the summers of 1929 and 1930 by the authorities of the Zoological Survey of India to Dumpep in the Khasi Hills, Assam, to carry out investigations on the ecology and bionomics of the anuran fauna of the torrential streams. During the second visit I was asked to make observations on and collect the early stages of the tadpoles of Rana afghana and of any species of Megophrys found in that area. I secured a fair series of the developmental stages of the tadpoles of R. afghana, but only mature tadpoles of two species of Megophrys. Dr Hora (1932) studied the tadpoles of R. afghana with special reference to the evolution and development of the adhesive apparatus. In view of the more or less complete developmental series of R. afghana tadpoles now available, Dr Hora very kindly suggested to me the investigation of the anatomy of the adhesive apparatus in this species.

In order to study the ecology of intertidal plants and animals it is desirable to choose a coast-line which is as simple as possible. When numerous factors vary simultaneously it is impossible to decide which of them is responsible for any change which may occur in the flora and fauna. The complexity of a shore which is ravined or boulder-strewn makes ecological work difficult. Unfortunately, certain major influences, such as wave action, involve a whole complex of subsidiary effects, so that the perfect locality, where each environmental factor would be found to vary in turn independently of all the other factors, cannot exist.

Raasay, one of the major islands of the Inner Hebrides, lies between the Isle of Skye, from which it is separated by the Sound of Raasay, on the west, and the mainland district of Applecross on the east. From Rudha na Cloiche, the southernmost promontory, to Rudha nan Sgarbh in the extreme north, the island has a length of over 13 miles, while the breadth from west to east varies from 3 miles near Ósgaig to a ½ mile just south of Arnish or at An Caol. The total surface area is about 27 square miles. The island is separated from Rona on the north by Caol Rònach, studded with islets and submerged rocks, a little more than a mile wide.

Twelve years ago Priestley and Ewing (1923) reported that in certain plants, normally showing but little development of stem-endodermis, an extensive formation of this layer could be induced by etiolation. (Note: In this paper the term endodermis is used only when the layer shows characteristic structural features—in the present case the Casparian strip.) A later paper by Priestley (1926) dealt with the same subject. The specified plants with which this result was obtained consisted of four closely related species, namely, Vicia Faba, V. sativa*, Pisum sativum, and Lens esculenta*, and also Solanum tuberosum. (The statements relating to the species marked with an asterisk are based on unpublished work carried out at Leeds, kindly placed at the author's disposal by Professor J. H. Priestley.) In these plants a primary endodermis, though present only at the base of the normal shoot, was described as extending to a considerable height in the etiolated shoot. It was concluded that the absence of endodermis from the greater part of the shoot of these plants, when grown under normal conditions, arises from the inoperation, in the presence of light, of the mechanism forming the Casparian strip.

The fossils which form the subject of this paper were collected in the summers of 1933 and 1934 by my son and myself at Earnstrey Hall Farm on the western slopes of Brown Clee in Shropshire. I had become interested in specimens collected at this locality by Mr W. Wickham King, M.Sc, which Sir Arthur Smith Woodward (1934, p. 566) named Corvaspis, and it was with the intention of obtaining further material of this genus that I got King to tell me exactly where he had found the best of his specimens. The section is in a small stream course 570 yards north 11° east of the farm and at the bottom of the escarpment formed by the “Psammosteus Limestone” and in the beds assigned to Zone I, 8 of King's classification (King, 1925). At the particular spot in question there is (in normal years) a waterfall over a 4-foot band of very tough calcareous pellet rock and conglomerate with occasional quartz pebbles up to about 1 inch diameter. This, especially its lowest few inches, had yielded the specimens that King had found, and from it we also collected abundant fragments of Corvaspis. Nearly all the specimens of Phialaspis (p. 439) which we discovered came from this rock, as did a fine example given by King to the British Museum and referred to later (p. 440). In 1934 the same rock yielded cf. Ctenaspis. Below this rock there is about 18 inches of grey shale, in which King had noted black streaks. This rests on several feet of red clay.

This research was begun in 1929, and a preliminary account of some of the results of the investigation was published in the Journal of Anatomy, January 1932. It was initiated by the discovery, during experiments on intravitam staining in the guinea-pig, that dye-bearing cells collected in large numbers in the endometrium of the uterus and that their incidence there was cyclic. The first results were obtained in animals injected immediately after parturition and killed at intervals thereafter. In a short series of virgin animals, the series being established by determining the date of heat by the vaginal smear method (Stockard and Papanicolaou, 1917, 1919), practically identical results were obtained.

Investigations have been made into the conditions of sedimentation of the Ordovician rocks of the Girvan district, with special attention to a series of greywackes, the Ardwell Beds, which are remarkable for frequent intraformational disturbances and unconformities.

In presenting the evidence it is proposed to follow the stratigraphical classification of Peach and Horne (1899), who, after careful consideration, adopted Lapworth's (1882) succession, except for a few minor alterations. These authors pointed out that the Llandeilo rocks north of the River Stinchar consist of Kirkland Conglomerate, Stinchar Limestone Group, and Benan Conglomerate, while south of the river they are represented by Tappins Greywackes and Glen App Conglomerate (see table p. 488).

St Kilda (Pl. V), the largest of a group of four islands located some 50 miles west of Harris in Latitude 57° 49′ N., Longitude 8° 34′ W., has an extreme length and breadth of 2½ by 1¾ miles and a circumference of about 8¾ miles. The highly indented coast-line is guarded by almost impregnable cliffs except in the Village Bay or Loch Hirta where there is a strip of sand a quarter of a mile long backed by a great storm-beach. The two principal inlets of the sea, Village Bay or Loch Hirta, and the Glen Bay open to the south-east and north-west respectively.

The genus Lyginorachis was proposed by Kidston (1923, pp. 18–19) to accommodate isolated petioles which, whilst their structure was similar in some respects to that in the petiole of Lyginopteris, could not be finally identified with the latter genus. Two species of Lyginorachis have been described (Scott, 1923, pp. 57-60; Crookall, 1931, pp. 27-34, pls. i-iii). Both came from Scottish Lower Carboniferous rocks; the one, Lyginorachis papilio Kidston and Scott, from the Cementstone Group (Calciferous Sandstone Series) at Norham Bridge on the Tweed, and the other, Lyginorachis Taitiana Kidston and Crookall, from the Auldton Limestone (Carboniferous Limestone Series) near Lesmahagow, Lanarkshire. Crookall also described a smaller and simpler petiole (loc. cit., pl. ii, fig. 3) as Lyginorachis sp., from the Cementstone Group at the Langton Burn, near Duns, Berwickshire.