noticeable as a possible Druidical rock basin, or as a socket for a central pole from which would radiate the roof-structure. On the ledge of wall near the entrance was another bowl resembling a "holy water font." (On Goonhilly Down, near the Lizard, found in a hedge hard by a rude cromlech (3 stones) were two basins, called by the folk the "bowl and ladle," the why and the wherefore profoundly hidden from the enquirer). These structures are surrounded by numerous "gurgos," broken-down fences or dykes, forming enclosures of fantastic shapes. The hut-dwellers baked their bread on "gradles "—our girdles-and venison roasted on flaming fern must have been "gey guid." ""*

WHITE HORSES.

These monuments of British history, or rather Saxon history, are found all along the line of the chalk downs. They are held to have been cut in commemoration of the exploits of Alfred, circa 879, near Ashdown, and one was known to be existing in 1100. Similar figures were cut at Cherhill, Alton, and Bratton, the latter probably in commemoration of Alfred's victory at Clayhill (A'glea). The "Scouring of the Horse" is well told by Hughes, and at various times the other white horses have been cut or rejuvenated, giving the critic the chance to sneer at those modernities. But as Mr. Wise remarks, "It is worthy of remark whether their authors had not preserved the tradition of some older monuments now obliterated, or of some older festival now forgot."

The myth of the "White Horse" is of great antiquity, and Mr. Thoms considers that memorials of the conversion of the Saxons to Christianity were preserved in form of white horses in ash groves. (Plenderleath, Wilts. Arch. Mag., Vol. XIV.).

Long may the "white horses" race along our shores, and, as British monuments, guard and defend our native land.

*See description of feast by Arnold in "Iyvinda."

NOTE.-Most of these historical notes were published by the present writer

in "Historic Wilts."

XXIII.--Note upon Crystals of Grossularite from
Corsiehill Quarry.

By S. J. SHAND, B.Sc., Ph.D., the Royal Scottish Museum, Edinburgh.

(Read 9th January, 1908.)

Two small specimens of garnet-bearing rock from Corsiehill Quarry were sent to the Royal Scottish Museum by Mr. Rodger in 1906. In examining these I noticed some interesting points connected with their genesis and crystalline form, and on my mentioning this to Mr. Rodger, he was kind enough to place at my disposal for investigation all the remaining material from this locality. I also made two visits to the quarry in his company, but failed to find any more of the material, which appears to have been exhausted.

The specimens were found by Mr. A. Grant Ogilvie and Mr. A. Gray, at the junction between the andesite and the large Tertiary basalt dyke which cuts the latter; they appear to have lain on the andesite (South) side of the junction, and to have been broken away during the quarrying of the basalt. I have had microscopical sections made of both the andesite and the basalt from various points in the quarry, but in no case is garnet present as a constituent of either rock. Even a specimen with visible garnets on the surface contained none internally.

The garnets are associated with incrustations of saponite, quartz, and calcite. In a few cases the garnets are seen to lie directly upon the andesite, but in general the following order of succession obtains among these four minerals, from within outwards :

(1). Layer of dark green saponite.

(2).

(3) and (4).

Layer of prismatic quartz.

Scattered crystals and grains of calcite, and
Scattered crystals of garnet.

The saponite forms a dark green to almost black layer on the surface of the decomposed andesite. The name saponite is used here in generic rather than specific significance, as the material is too earthy and impure to admit of exact determination; it is probable that it contains more than one member of the saponite family.

Upon the saponite there is in most cases a layer of compact crystalline quartz consisting of prisms standing perpendicular to the surface of the saponite, and terminated above by the usual hexagonal pyramids. The surface of this quartz layer is stained brown, and it shows here and there traces of a second deposition of saponite.

The calcite is in roughly-shaped scalenohedral crystals and rounded grains. These are sprinkled liberally over the quartz, and among and upon them lie the garnets. The calcite and the garnets appear

to have been deposited almost simultaneously, as each species may be observed to cap crystals of the other.

The garnets belong to the species grossularite or calcium-aluminium garnet, and to the variety known as cinnamon-stone. They vary in colour from nearly white to pale yellow and rich cinnamon-brown. In size they are minute, few exceeding 1 millimetre in diameter and none rising to 2 millimetres. The best individuals are beautifully transparent, and possess a high lustre, thanks to which their minute facets readily catch the eye. Faces belonging to at least five forms are recognisable; they are thus by far the most richly facetted garnets yet found in Scotland. (See Fig. 1).

The predominant form is the icositetrahedron n = (211). The faces of this form are large and bright. They show numerous fine striations parallel to the edge dn (Fig. 2), which are caused by the presence upon n of a vicinal hexakisoctahedron. The faces of the latter are sometimes developed as indicated in the lower half of Fig. 2, each face of n being replaced by two faces which lie almost, but not quite, in the same plane. The above type of striation is quite unusual in garnet.

=

The rhombic dodecahedron d (101) is represented by very small facets, of which the full complement is seldom present. They show no peculiarity.

A second icositetrahedron is represented by four minute and strongly curved faces (m) at the octahedral corners. The curvature of these faces is so pronounced that the angle nm varies from a minimum of about 10° to a maximumn of nearly 20°. The former value agrees with that required by the form m = (311), to which these faces are here referred. This form has not previously been recognised in Scotland.

Truncating the edges nn are the narrow facets > = (332) of a triakisoctahedron. These are always rounded and not adapted to measurement, but the formula is given by their position. This form is also new to Scotland.

The remaining form is a hexakis-octahedron indicated in Fig. 1 by the letter. This replaces each octahedral edge by two facets, and lies therefore in the zone nn, and not in dn as is usually the case. Its faces are sharp and brilliant, but so narrow that the strong diffraction caused by them in reflected light prevented the formation of an image suitable for accurate measurement. From their zonerelations it is deducible that the formula is of the type (2 h. h. 1). By approximate adjustment of each face to its position of brightest reflection, I obtained the following rough values for the angles :nat, 91°; t ▲ t, 28°; t a n, 91°.

Since the above was written I have been enabled, by the kindness of

Professor Lewis and Dr. Hutchinson, to repeat these measurements with a small Fuess goniometer in the Mineralogical Laboratory at Cambridge. For the crystal registered in this Museum under the number [370-179a] I obtain the following values, measuring from n to n across an octahedral edge :

If = (14 7 4), the theoretical values of these angles are n at = 9° 46'; ta t = 28° 39'

The agreement with this form is therefore fairly good. The form (14.7.4) is entirely new for garnet, the usual hexakisoctahedra lying in the zone dn.

From the nature of the occurrence and of the associated minerals, there can be no doubt whatever that these garnets are of aqueous origin. As they are found only along the line of junction of the andesite with the intruded basalt, it is reasonable to conclude that they date from the time of irruption of the latter, and that they owe their formation to the action of the heated waters which accompany every demonstration of igneous activity. The curvature of the faces appears to be an original phenomenon, as it is in many diamond crystals, and not a result of secondary re-solution.

XXIV.-Notes on the Natural History, Geology, and Antiquities of Duror, Argyllshire.

By REV. G. A. FRANK KNIGHT, M.A., F.R.S.E.

(Read 13th February, 1908.)

It is often a question by no means easy to solve where one's holidays should be spent. There are some places so crowded with visitors, so given over to noise and bustle, that even although the air may be bracing, and the scenery and surroundings may be pleasing, the tired city dweller returns to his work at the close of his holiday with the sense that he has never been away from the roar of the world's din. To a lover of quiet glens, of lonely shores, of wide silent moors, of sweet rustic life, such fashionable centres offer no charms. But if anyone desires a locality with beauty in whatsoever direction one looks, with delightful seclusion from "the madding crowd," with ample scope for scientific and archæological study, with the purest air and the grandest mountain and loch scenery, I can heartily commend Duror in Argyllshire.

Anyone who remembers the character of the weather last August will be ready to confess that a wetter holiday month could hardly be imagined. As a matter of fact, it was our privilege to enjoy only two days during which it did not rain. The incessant torrents to a large extent prevented my studying the district as I would have liked, and I must make the wretched weather my apology for the meagreness of the following notes. on the locality, and for the unsatisfactory character of the slides which I show on the screen. In such sunless and dripping weather it was almost impossible to obtain good negatives. I propose to take up (1) The Marine Conchology, (2) The Geology, and (3) The Antiquities of Duror.

I. THE MARINE CONCHOLOGY.

Not having my dredge with me (fortunately-for the weather all the month would have been too stormy to use it), I had to be content with shore collecting. The mollusca of the district, however, are not numerous. The shore is as a rule rocky, and even where bays do occur, the beach is shingle of various degrees of roughness, and the patches of sand are few and far between. Loch Linnhe, moreover, penetrates so far into the heart of the country that it is not in close touch with the main ocean, and its waters therefore are not so prolific in mollusca as those swept daily by the full strength of