The accompanying map is the outcome of a survey of Shetland, conducted during part of several summers.

Almost the whole of the junctions laid down were visited by myself ; for the few that were not, and also for certain of the lines continued inland—which from an almost unbroken covering of peat and turf, are more or less conjectural—I am indebted to Dr. Hibbert's narrative.

To Professor Geikie's map of Scotland I am indebted for the southern limit of the isolated mass of sandstone of the Lerwick district.

Anyone who contemplates on a map the great size of this island in geographic extension, and its extraordinarily irregular outline, would conclude that it must be composed of various formations, and that the connections and dispositions of these must be intricate in the extreme.

The latter conclusion will, however, upon a more careful observation, be found to apply only to the western portion of the island. However deeply trenched by the sea iu its lower grounds, however persistently thrusting its cliffs oceanward in its higher, the ridged and furrowed outlines of the eastern, disposed in almost rectilinear sequence, indicate but one formation,—various as the members of that formation may in mere lithological features appear.

Referring to my notes in number 3 of the Mineralogical Magazine for 1877, I observe one or two little matters which I hoped the march of time would improve. One desideratum, a good hotel, the Hotel d'Anniviers, has been established at Vissoie, and M. Monnet, the landlord, is a man of intelligence, and, without pretending to scientific lore, has much information concerning the mining, geology, botany and history of the neigbbouring valleys. Jean Martin, my guide in many a tough scramble, has married, and established a small restaurant with a couple of bedrooms at Vissoie, which place may now fairly be said to have become a centre for many interesting excursions ; as, to the Bella Tola, whence the panorama is very grand,—to the Cols leading from the Grimenz Thal to the Val de l' Arolla and Evolena, and also to the Turtman Thal, and so across to St. Niklaus in the Zermatt valley.

The difficulty of devising a natural classification of minerals has been recognized by most mineralogical writers, and up to the present time no very satisfactory arrangement has been generally adopted. In the classification of the animal and vegetable kingdoms, although the systems have greatly varied from time to time, something like uniformity has pretty generally existed ever since Linnaeus proposed his famous system of double nomenclature ; but in mineralogy every writer does that which is right in his own eyes—not because there is no king—but because they are many and mutually antagonistic.

In the following short paper I propose to describe what I have lately done in improving the method of studying the optical characters of the minerals which I treated at greater length in my address last year at annual meeting at Plymouth.

It is a curious example of how a method may be invented, and then lost sight of, that the determination of the index of refraction in the way there described was proposed by a French savant upwards of a hundred years ago. I have not yet consulted the original publication, but I very strongly suspect that the proposal was more theoretical than practical, and that with the instruments then at disposal the results were found to be so inexact that the whole system became obsolete and practically forgotten.

Rather more than twelve months ago Sir D. Jephson Norreys, Bart., of Mallow Castle, county Cork, directed my attention to a beautiful mineral which had been obtained by a friend and neighbour of his, Miss Cotter, of Rockforest. Through his kindness I was enabled to see the circumstances under which the mineral occurred ; and also to obtain from Rockforest some fine specimens thereof.

The mineral itself is specially characterised by the peculiarity of its lustre, which it possesses in a high degree, producing an aspect of a very unique character. Concerning the kind of lustre, this is somewhat difficult to define.

The Iron Crystals described by Mr. Beardsley in the present volume of the Mineralogical Magazine (p. 223) exhibit the remarkable structure illustrated in figs. 1 to 5, plate VII. Fig. 1 represents the crystals drawn natural size, and fig. 2 is a section of one of them in a plane parallel to a face of the cube—shewing the structure developed by etching. Figs 3 and 4 illustrate other modificatious of structure which have been observed in iron crystals, and for which I am indebted to Dr. C. O. Trechmann of Hartlepool, and fig 5 may perhaps indicate the ultimate molecular structure.

This lies to the south of Unst ; it is interesting as forming the connecting link which unites the rock masses of Fetlar with those of Unst.

Its western extremity is composed of diallagic rocks, containing crystals of that mineral (Hb) very inferior to those of Balta : its middle portion consists of various rocks of the "chloritic series" ; towards the east there is seen a bed of graphitic slate (H) quite similar to, and doubtless the continuation of one which appears near Trista V oe in Fetlar.

Some extremely beautiful but rather soft gems of various green hues have been extensively sold during the last year or two. They are found in the gold washings of the river Bobrowska, near Poldnewaja, district of Syssersk, in the Urals, and occur in the form of nodular masses, varying in size from a pea to a chestnut. These closely resemble the jelletite from Monte Rosa. All the specimens which I have examined present a roughly radiate structure, are irregularly fissured and veined, and are coated with a talcose mineral, glistening and nearly white. The exposed uninjured surfaces of the nodules are minutely and unevenly striated. The mineral shows no cleavages, but the broken surfaces have a resinous or semi-adamantine lustre, and show a small conchoidal fracture.

In a former paper I pointed out that it was likely that some of the rocks and minerals at present classed as magnetic were so only by reason of intermixture of a silicate of iron having the formula of Fe 3 O4 2 Si O2, or Fe8 Si2 O8, and that when purified from the particles of this body by pulverization and magnetic sifting, they might be found to be either much less magnetic or even diamagnetic.

In the course of an examination of such Scottish garnets as were pure enough for analysis, I found that four had a composition which was very similar, but which differed from that of any known, and that to an extent entitling them to be regarded as constituting a well-marked variety.

These garnets, from their purity of colour and transparency, would all be classed as precious garnets–they are, indeed, the most brilliantly coloured which the northern portion of the kingdom affords.

In my address at the Annual Meeting of the Mineralogical Society last year, in Plymouth, I gave an account of a new method of studying the optical properties of minerals, and of determining their indices of refraction. I then showed that an approximate knowledge of these characters would, in many cases, enable us to identify minerals with great confidence, and also would sometimes furnish us with valuable hints as to their general chemical composition. My remarks had reference exclusively to detached specimens of more or less considerable thickness, and I scarcely thought that it would ever be possible to apply this method successfully with comparatively high powers in the case of thin sections of rocks. I feared that certain errors, which appeared to be unavoidable, would then become so great in proportion to the difference in the indices of many of the commoner minerals, that it would be impossible to distinguish them with confidence.