The genus Desorella was founded by Cotteau in 1855 to include four species from the Corallian named D. icaunensis, D. Orbignyana, D. elata, and D. Drogiaca, and one from the Neocomian—D. incisa. The first and the last of these five species were subsequently shown to belong to the genus Pyrina, the second was placed by Etallon in a new genus, Pseudodesorella, whilst the fourth has been referred to Hyboclypeus. So that of the five species originally regarded as belonging to Desorella, one only, D. elata, is left in that genus. Another species, described in 1862 as Desorella Guerangeri, was afterwards shown to be a Pyrina. Cotteau accepted these modifications, and, in 1873, gave an emended diagnosis of the genus, naming as the type Desorella elata, and describing another species, D. Grasi. M. Jules Lambert is of the opinion that Pachyclypeus semiglobus (Goldftiss) should be placed in the genus Desorella.

In a notice published on page 160 of the present volume it was suggested that Arsinoitherium, of which a brief description was there given, must be placed in the order Amblypoda, constituting a new family, the Arsinoitheriidæ. Having further considered the evidence available, and having, moreover, had the advantage of some discussion on the subject with Professors H. F. Osborn and W. B. Scott, I have now come to the conclusion that Arsinoitherium differs from the Amblypoda in so many points that it seems necessary to refer that remarkable mammal to a new subdivision of the Ungulata of equal value with the Amblypoda and Probosoidea, to both of which a certain degree of relationship may exist. For this new order the name Barypoda is proposed, in allusion to the massive character of the limbs in the species at present known. The existence of two orders so marked as the Proboscidea and the Hyracoidea in the Eocene beds of Egypt, and their absence at that period from the rest of the world (with the possible exception of South America), at least make it seem likely that in an area so isolated there were other equally distinct groups which died out before circumstances became favoui'able to allowing them to pass over to other regions. That Arsinoitherium may be the representative of such an order is probable, for its great size and highly specialised character point to its being the closing member of a long line, of which the earlier forms are at present quite unknown, and must be sought in earlier horizons in the Ethiopian region.

On hearing that Mr. G. E. Dibley, F.G.S., had exhibited at the June meeting of the Geologists' Association some test-plates of Marsupites from the Chalk of a new road near Russell Hill, I made inquiries of him, and he most kindly told me the particular locality where he and others had obtained these fossils. As the place was within an easy walk of my home, I visited it, in company with the younger members of my family, on an evening in June, and found, as Mr. Dibley had told me, that the greater part of the Chalk had been refilled into the sewer-trench, and the residue had been spread over the roadway and was now partly trodden down by the traffic. We patiently broke a number of the remaining blocks of chalk, and, in spite of the fact that Mr. Dibley had already carefully worked at the place, succeeded in finding not only a couple of plates of Marsupites with Echinocorys scutatus, etc., but also some smaller inconspicuous plates which, when cleaned from the matrix, proved to be test-plates of the unstalked, free-swimming crinoid, Uintacrinus, Grinnell. These showed the existence at this place of the lower portion of the Marsupites-zone, which has been designated by Dr. Rowe the “Band of Uintacrinus.”

During the last fifteen years slight earthquakes have occurred in Cornwall on eight occasions, the dates being Oct. 7, 1889; Mar. 26, 1891; May 16 and 17, 1892; Jan. 26, 1896; and Mar. 29, April 1 and 2, 1898. The Pembroke earthquakes of Aug. 18, 1892 (0.24 and 1.40 a.m.), and Nov. 2, 1893, and the Hereford earthquake of Dec. 17, 1896, were also felt in the county. Local earth-shakes, probably connected with mining operations, occur occasionally, as on June 4 and 10, 1902. Under the same heading should perhaps be included the shock of Aug. 27, 1895, near Blisland, which I was led to class as seismic on account of its very elongated, though small, disturbed area.

The valley of the Tochi River is an outlying corner of the British Empire in India forming a portion of Waziristan, the boundary of which was delineated in 1894–5 by an Anglo-Afghan Commission from the Afghan provinces of Khost on the north and Birmul on the west. Mr. F. H. Smith, of the Geological Survey of India, accompanied this Commission as geologist, and his observations “On the Geology of the Tochi Valley” were published in 1895 in the “Records of the Geological Survey of India” (vol. xxxviii, pt. 3, pp. 106–110, pl. iii). On p. 109 he says:—“The range of hills between Idak and Mirán Shah is formed by an anticlinal ridge which approximately strikes north and south, and which is composed of these lower eocene beds. In the core of the anticlinal a considerable thickness of massive dark grey limestone is exposed, in which I could find no fossil remains; the age of this limestone is therefore doubtful, and there is no evidence of any kind to show whether it belongs to the lowest tertiary or upper mesozoic age.”

The experimental work of this investigation was done by Grace D. Bradshaw in the chemical laboratory of Cornell College. The purpose was to determine whether the silica exists in a free condition or is in the form of a silicate; also to ascertain whether the iron is in the ferrous condition as carbonate or is in the form of ferric oxide. The rocks abound in many parts of Iowa, and belong to the Niagara formation. The stratified character even in a small section is apparent, and the layers differ somewhat in composition, as shown by the varying amounts of iron visible in different portions. The rocks are used as building stone to manufacture quicklime and in MacAdam paving.

I Used to think that the corrugations of the earth's crust were due to compression through the shrinking of the interior. To judge of the sufficiency of this cause the first thing to be done is to seek a measure of the compression, and then to compare the result of the effects of cooling with the actual amount of compression. The most satisfactory measure appears to be the thickness of the layer which the corrugations would form if levelled down. The question then becomes one of how much. In 1863 Lord Kelvin (then Sir W. Thomson) formulated a law of secular cooling upon the hypothesis that the interior is solid. Adopting a probable value for the contraction of rocks in cooling, I calculated the thickness of the layer which would be produced by the corrugations resulting, and found it far short of that which the existing inequalities would form if levelled down. Mr. Mellard Reade and Dr. Davison subsequently discovered the existence of a level of no strain within the crust, and this greatly reduces the possible amount of corrugations. The conclusion at which I arrived was that, on the hypothesis of a solid globe, secular contraction through cooling would not account for the corrugations.

The author refers to his work in former years among the High-level Plateau Gravels south of the Thames, chiefly in Berks and Surrey, the results of which were given in various papers from ten to twenty years ago.6 The present note may serve as a supplement to those papers, in which the conclusion was arrived at that the gravels in question were to be regarded as distinctly of riverine origin and, upon the whole, of Pliocene age.

During a visit to Mount Sorrel at the end of last year I collected a number of specimens of the dark-coloured patches which are so common in the granite. I have had about a dozen of these sliced, and an examination of them has yielded some results which seem worthy of a brief description.

This communication brings up to date the record of losses on the _L Suffolk coast, and continues the report presented at the Ipswich Meeting, 1895, of which details were published in the Geological Magazine for January, 1896. Since that time scarcely a year has passed without the winter gales and high tides doing mischief at one or more points of the coast embraced within the above-mentioned limits; but whilst Lowestoft and Pakefield, Covehithe and Baston have all suffered very considerably, the cliffs at Dunwich remained until quite recently almost unaffected.

The bare and rocky hill known as Holyhead Mountain is of considerable interest in connection with recent geological events, standing as it does some thirty miles out from the highlands of Carnarvonshire into the Irish Sea Basin; and in such remarkable isolation, for it is much the highest of the five hills which rise above the general level of the platform of Anglesey.

Its height is only 721 feet, but so strongly featured is it, especially towards the west, that one feels the term ‘mountain’ to be no misnomer, and can hardly believe it to be really lower than many of our smooth wolds and downs of Oolite and Chalk. Being composed, moreover, of white quartzite (or more properly of quartzite-schist), and being so bare of vegetation, it recalls much more vividly certain types of scenery in the Scottish Highlands than anything in those Welsh mountains that one sees from its sides. Towards the east it slopes at a moderate angle, but a little west of the summit it is traversed by a very strong feature, due to a fault, running nearly north and south, along which is a line of great crags, facing west, and prolonged northwards into the still greater sea cliffs towards the North Stack. Beyond this the land still remains high, but is smoother in outline, a somewhat softer series of rocks extending from the fault to the South Stack, where the high moors end off in great cliffs above the sea.