The geology of the Bermuda group of islands has been so exhaustively treated by Professor Verrill, Dr. Alexander Agassiz, and Lieutenant (afterwards General) Nelson (in what will always be the classic memoir on the subject) that it may seem presumptuous for another description to appear. However, as there are some new facts to record, and nothing has yet appeared in the GeologicalMagazine on the geology of Bermuda, a paper may be not altogether out of place.

In a previous paper published in this Magazine, I gave reasons for the belief that neither the Gulf of Suez nor the valley of the Nile owes its origin to trough-faulting, as was at that time generally supposed. The skeleton of the argument was as follows:—

1. Extensive faulting at the faces of the scarps of the Nile Valley and Gulf of Suez cannot be held to be evidence of trough-faulting, since the same can be observed along the scarps of the Wadi Araba, the structure of whose floor shows it to be an eroded anticline, and the faults along its scarp-faces to be merely huge landslips.

2. The faulting observable immediately along the coast of the Gulf of Suez is of an exactly similar nature to that above mentioned, and along part at least of the gulf the strata dip away from the sea on the opposite coasts, leading to the inference that the Gulf of Suez is, like the Wadi Araba, an eroded anticline.

3. If it be granted that the Gulf of Suez is not a trough-fault, the argument against the Nile Valley being a trough-fault is strengthened, the support of parallelism being removed.

During one of the excursions arranged by the International Geological Congress held in Sweden, 1910, we visited, under the leadership of Professor Högbom, a group of islands skirting the shore north of Sundsvall, on the Gulf of Bothnia. Among these islands Alnö and Långörsholmen are of particular interest on account of a complex of alkali igneous rocks forming the whole of the latter and the northern corner of the former. It is to this small but highly interesting group of rocks that I wish to call the attention of the reader, since apart from the short but convincing ‘guide’ prepared by Professor Högbom (1), scarcely anything has been written on the subject in the English language. The expedition was very skilfully planned, and we were thus enabled to visit several critical exposures and collect numerous specimens for microscopic study.

In April, 1879, William Davies published in the GeologicalMagazine (n.s., Dec. II, Vol. VI. pp. 145–8) a paper on “Some Fish Exuviæ from the Chalk, generally referred to Dercetis elongatus, Ag.; and on a new species of Fossil Annelide, Terebella Lewesiensis”. Herein he discussed certain longitudinal or tubular agglomerations of fish-debris, originally named Murœna(?) Lewesiensis, by G. A. Mantell, but subsequently assigned by L. Agassiz to the fish described by him as Dercetis elongatus. References to the various papers and books in which these remains had been mentioned will be found in the paper quoted. From his profound knowledge of fossil fish Davies was able to show that these agglomerations contained the remains of more than one species of fish, and he considered that the fragments had been collected and affixed to their tubes by annelids allied to the modern Terebella. It is well known that some living species of that genus have similar tube-building habits, so that the suggestion made by Davies has been generally accepted, and the specimens in the British Museum on which he based his conclusions have since then been labelled Terebella(?) Lewesiensis Mantell sp.

Description. Body elongated, narrow, composed of a longitudinal series of eight imbricating plates, and marginal girdle (?) of spicules.

Limb-bones.—Since the humerus is known in Hylœsaurus (?), Omosaurus, and Stegosaurus, it is quite easy to compare the humerus of our new species with that of the other genera mentioned. Whether the shaft is hollow as in Omosaurus or solid as in Stegosaurus, cannot be ascertained without breaking the specimen; the outline of the new humerus is, however, that of Omosaurus, not that of Stegosaurus.

If we take a rapid glance over the volcanic history of the British Isles we find that eruptions having the spilitic facies have occurred repeatedly, over a wide area and on a large scale.

A recent paper by Mr. A. R. Horwood on “Aragonite in the y Middle Lias” raises again the question of the condition of the calcium carbonate in fossil shells. Mr. Horwood speaks (p. 176) of Jurassic shells as being “preserved in aragonite”, and quotes Dr. Sorby as being of the same opinion. But is Mr. Horwood sure that his shells are still in the aragonite condition? Sorby, when he writes of “aragonite shells”, means as a rule shells that were formed of aragonite at the time of the death of the originating animal; but he describes how, in many of our fossiliferous rocks, such shells are represented either by hollow spaces or by pseudomorphs in granular caleite. He points out that oolitic grains, originally consisting of aragonite, have similarly been changed into calcite, the more stable form of calcium carbonate. Some oolitic grains, however, were thought by him to have consisted of calcite from the outset. Doelter, after summing up Linck's researches on calcite and aragonite, states that calcareous oolites now consist of calcite; and we have no doubt that the “recrystallized aragonite concretions”, quoted by Mr. Horwood from Sorby, were held by Sorby to be at the time of examination in the calcite state.

The preglacial platform of marine erosion discussed in the following pages was first discovered in the Island of Colonsay in the course of the geological survey of that island carried out in the summer of 1907, and the information then acquired is communicated with the permission of the Director of the Survey. The subsequent tracing of the shoreline in the adjoining islands was effected with the aid of a Government grant for scientific research in the gift of the Royal Society of London. For clearness and convenience of reference the subject will be treated under the following heads:—

I. Introductory remarks indicating the aims of the investigation.

II. Colonsay and Oronsay—description of the preglacial platform.

III. Islay.

IV. Mull and Iona.

V. The Treshnish Islands.

VI. Conclusion, setting forth the present state of the subject.

(4) Recent Æolian Sands.—On the south shore the sand has made or is making some encroachments on the land, in others on the sea. The ‘live’ sand-dunes are principally to be found at Tuckerstown, but there are some movements at Elbow (or Middleton) Bay and at Warwick Long Bay. The shell-sand is mostly made up of the broken shells of various marine species, some few being entire, with which are mingled the broken tubes of serpulæ, fragments of nullipores and corallines, and the red shells of a Vermetus, viz. Teganodus (Siliquaria) ruber, Schum. This latter is so plentiful as to give a distinctly red appearance to the sand of some beaches.

The second fundamental problem of petrogenesis is the following question:—By what processes have there been derived, from the original magma or magmas, all other magmas known as igneous rocks? It is now almost universally admitted that many igneous rocks are genetically connected and produced from an original magma by differentiation, and that in many cases differentiation may be considered as liquation or separating into secondary magmas. But is differentiation (Spaltungen) alone sufficient to explain the formation of all igneous rocks? And by what is a magma stimulated to differentiation? These are two questions which must be elucidated. It must be first of all emphasized that not every magma is subject to differentiation. My standpoint is that the principal factors producing and regulating differentiation are on one side the process of crystallization (differentiation by crystallization, Krystallizations-differenzierung), and on the other the chemical composition of the magma and its tendency to the formation of eutectics (magmatic differentiation).

Remains of Shrews have long been known to occur in the Norfolk ‘Forest Bed’, and have been discovered in several British Pleistocene deposits. Hitherto they have been referred to one or other of the three species at present inhabiting this country, but having had occasion lately to examine nearly all the available material, comprising representative series of specimens from each of the known horizons, I find that it is not until we reach the latest Pleistocene deposits that we meet with, remains of species indistinguishable, with the material before us, from the living British forms.

The specimen represented in the accompanying figure was obtained I from the Coal-measures of Westhoughton, near Bolton, Lancashire, and has been kindly lent to me by Mr. Thos. Midgley, of the Chadwiek Museum, Bolton. It is the nearly complete abdomen (opisthosoma) of an arachnid of the order Anthracomarti. The concavity of the outer margins of the pleural laminæ, producing the scalloped outline shown in the figure, indicates at once that the specimen must be assigned to the family Brachypygidæ; but it is less obvious to what genus it may belong. Mr. Pocock's monograph (see preceding footnote) gives only two genera as known in the family, namely Brachypyge, Woodward, and Maiocercus, Pocock. Comparison between these two is rendered difficult by the fact that of the former only the dorsal surface is known, of the latter only the ventral; of neither has any part but the opisthosoma yet been recognized. Accordingly, in differentiating his genus Maiocercus from Brachypyge, Mr. Pocock was compelled to fall back upon two characters—the general shape of the opisthosoma and the angles made by the anterior pleural laminæ. Brachypyge is defined as having the “opisthosoma much longer than wide; pleural laminæ of the second and third pleura-bearing terga inclined slightly backwards” Maiocercus as having the “opisthosoma much wider than long; pleural laminæ of the first, second, third, and fourth sterna inclined slightly forwards”.

It is a noteworthy feature of the most ancient crystalline limestones that they are generally dolomitic, or at any rate frequently so. They often consist of practically pure dolomite; whilst associated with this in some cases is pure crystalline magnesite, which is very liable to be mistaken for saccharoidal dolomite.

The single specimen which forms the subject of this note was sent to me for examination by Mr. Hewitt, of the Albany Museum, Grahamstown, Cape Colony; it was found at Cyphergat and presented to the Museum by Mr. S. R. Gardner. Fossil plants from Cyphergat in Cape Colony have been described by Feistmantel and, more recently, by myself from beds containing the remains of a flora assigned to the Rhætic age. Our knowledge of this flora is based on material obtained from the Molteno Beds at the base of the Stormberg Series at Molteno, Stormberg, Cyphergat, and other places.

With the question of conformity or unconformity between the Lower Tertiary and Upper Cretaceous, there is associated a problem of great economic importance to New Zealand. The subject involves something more than a mere academic discussion. If conformity exists, then we have only one coal-bearing formation; but if unconformity, then we have two. I believe that both the stratigraphical and palæontological evidence is overwhelmingly in favour of the latter.

Since the Omosaurus of the Kimeridge Clay may still be regarded as the only well-known European representative of the Stegosauridæ, it seemed advisable, after discussing in previous papers the Ornithopodous Hypsilophodon and the Acanthopholidid Polacanthus, to examine a representative of this type. I am therefore greatly indebted to Dr. A. S. Woodward for permitting me to do so at the Natural History Museum, and also for putting at my disposal a magnificent hitherto undescribed Stegosaurian discovered by Mr. Alfred Leeds, F.G.S., in the Oxford Clay of Fletton, near Peterborough.

It may be taken that the general tendency of the evolution of surface ornament (and therefore of the radioles) in the Holectypoida was a dual one. The largo tubercles become gradually restricted to the adoral surface, and often reduced in number, while the actual density of the tuberculation increases. This increase is attained, either by a simple multiplication of the normal vertical series, as in Discoïdea (this being sometimes associated with the transformation of granules to tubercles at the ambitus), or by a more uniform distribution of those series over the area of the plates, and the frequent substitution of two tubercles in the place of one, as in Conulus. Both these types of change lead away from the primitive, ‘Regular’ condition of tuberculation, and trend towards the Clypeastroid and Cassidulid arrangement.

In the early numbers of the Geological Magazine the ‘Dragon-tree’ received considerable attention. It was thought to be a Monocotyledon and described as a species of Dracæna by Mackie in 1862 (Geologist, p. 401, pl. xxii). The specimen appears to have attracted an unusual amount of interest prior to 1870, and it is referred to by Mantell and other contemporary writers. Carruthers placed it among the British fossil Pandanaceæ in 1868 (GEOL. MAG., p. 154). Like other specimens from the sandstones of the Iguanodon Quarry (Lower Greensand of Maidstone), no internal structure appeared to be preserved, and the determination of its nature rested entirely on the external features.

Among the Crustacea from the Coal-measures of Coseley, near Dudley, which have been lent to me for examination by Mr. Walter Egginton, are a number of specimens of a form probably identical with that described by Dr. Henry Woodward as Prœanaspides prœcursor. These specimens make it possible to supplement, in some important particulars, the original account of the species, and throw fresh light on its relations to the living Syncarida. Reasons are given below for regarding the genus Prœanaspides as a synonym of Palœocaris, Meek & Worthen.