d'arbres innombrables, les cellules pyramidales, qui, grâce à une culture intelligente, peuvent multiplier leurs branches, enfoncer plus loin leurs racines, et produire des fleurs et des fruits chaque fois plus variés et exquis.

Du reste nous sommes très loin de croire que l'hypothèse que nous venons d'esquisser puisse à elle seule expliquer les grandes différences quantitatives et qualitatives que présente le travail cérébral chez les divers animaux et dans la même espèce animale. La morphologie de la cellule pyramidale n'est qu'une des conditions anatomiques de la pensée. Or cette morphologie spéciale ne suffira jamais à nous expliquer les énormes différences qui existent au point de vue fonctionnel entre la cellule pyramidale d'un lapin et celle d'un homme, ainsi qu'entre la cellule pyramidale de l'écorce cérébrale et le corpuscule étoilé de la moelle ou du grand sympathique. Aussi à notre avis est-il très probable qu'en outre de la complexité de leurs rapports les cellules pyramidales possèdent encore une structure intraprotoplasmique toute spéciale, et même perfectionnée dans les intelligences d'élite, structure qui n'existerait pas dans les corpuscules de la moelle ou des ganglions.

"On Rocks and Minerals collected by Mr. W. M. Conway in the Karakoram Himalayas." By Professor T. G. BONNEY, D.Sc., F.R.S., and Miss C. A. RAISIN, B.Sc. Received February 15,-Read April 19, 1894.

During his journey in the Karakoram Himalayas, Mr. W. M. Conway collected more than 300 specimens of rocks and minerals, which, however, were generally rather small. These were sent to us for examination at University College, London. Thin slices have been prepared of the specimens which promised to be the more interesting. Of the rest, the mineral composition was verified in cases of doubt by examining pulverised fragments under the microscope.

Since the detailed results of our examination, which practically form an annotated catalogue of the specimens, will be printed as an appendix to Mr. Conway's forthcoming volume, we restrict ourselves in this paper to a summary of our work, and to a notice of a few specimens which appear to be of more than local interest.

(1.) General Description.

Commencing with the crystalline rocks, and with the most basic of these, we find one specimen of a dark green serpentine, containing

* Since this paper was read, the first or descriptive part of the work has been published (‘Climbing and Exploration in the Karakoram-Himalayas '), which gives the positions of the localities mentioned herein.

small glittering crystals of bastite, brought from débris at foot of a slope at the village of Mapnun on the Burzil Pass. The rock has been evidently affected by pressure, and is practically identical with a type of serpentine rather common in the Alps.

Many varieties of diorite have been collected. Of these, some are almost hornblendites (see below), others are normal diorites varying from coarse to fine grained, others, again, are really hornblendeschists. In some of the last it is likely that the foliation (as we believe to be generally the case with the hornblende-schists of the Lizard and of Sark)* is the result of fluxional movements anterior to consolidation, while in others this structure is more probably due to pressure, and to consequent mineral changes subsequent to the first solidification of the rock.

Granites are rather numerous. Some are of a normal type, moderately coarse grained, the mica (variable in quantity) being generally biotite. Certain of these are slightly gneissoid in structure. More definitely foliated, almost certainly as a result of pressure, are some rather micaceous (biotite) granites. Yet more distinctly gneissic rocks occur, with a mineral banding as well as a foliation, in which pressure modification is generally to be noted. One group of these has a rather markedly different character; they are fine grained gneisses, modified by pressure (to which, however, the texture does not appear to be due), not rich in quartz, consisting mainly of felspar (orthoclase or microcline and plagioclase) and biotite, with a more or less definitely banded structure. They present a considerable resemblance to certain rocks found in the district about Blair Athol (Scotland), which Dr. H. Hicks, in consequence of their rather peculiar aspect, has named the "pepper and salt" gneisses. We find also a coarse granite, rather porphyritic in structure, and rendered gneissoid by pressure, which contains large red impure garnets, up to about 2 in. in diameter. Similar garnets, but of smaller size, also occur in several varieties of granite and gneiss; one of these might almost be called a kinzigite, others are granulites (leptynites).

The compact acid igneous rocks are not numerous; but a rather remarkable series from the neighbourhood of the Golden Throne calls for a little notice. Some come from the moraine proceeding from the west foot of the mountain. These are compact, but show a schistose structure with slightly micaceous or talcose aspect, indicative of crushing. They are parti-coloured; a grey, varying from slightly to markedly greenish, being streaked, spotted, and blotched with a dull Indian red, small white specks showing in some of the patches. The texture of the lighter part under the microscope is * T. G. Bonney and General C. A. McMahon, Quart. Jl. Geol. Soc.,' vol. 47 (1891), p. 497; E. Hill and T. G. Bonney, ‘Quart. Jl. Geol. Soc.,' vol. 48 (1892) *p. 145.

speckled, or somewhat fibrous, with indications of minute aggregate polarisation, this being most conspicuous when it makes an angle of 45° with the vibration planes of the crossed nicols. The fibrous mineral very closely resembles the so-called sericite of some porphyroids. Here and there are grains of quartz, decomposed and sometimes broken felspar crystals, and some dark reddish, flaky patches. The dark microscopic spots contain corroded grains of quartz, decomposed or partly corroded felspar crystals, and the same dark red flaky minerals in a crypto- to micro-crystalline matrix. In one rounded lump, at first sight very like a pebble, some grains of crystalline calcite are enclosed. The smaller patches, though varying in detail, present a general resemblance, and there can be little doubt that they are fragments of a devitrified acid igneous rock. In the absence of definite characters, it is difficult to speak positively as to the nature of the matrix, but most probably it was once a volcanic glass, which has since undergone micro-mineralogical change, mainly in consequence of pressure.*

Though the red patches occasionally look very like pebbles, it is more probable that they are due to a flow brecciation. If we are right in our inferences, these rocks of the Golden Throne indicate an outpouring of acid lavas prior to the mountain making. Other specimens from the same district are generally similar to the abovedescribed, but present varietal differences, and a small specimen of an Indian red colour from débris on the Baltoro glacier is very probably a tuff, though the amount of crushing makes it difficult to be sure.

The crystalline schists include epidote-, Piedmontite-, and various mica-schists. The first of these very probably occurs as a band in a hornblende-schist, and a rock very similar to it may be found at more than one locality on the Lizard peninsula, in Cornwall. The second schist will be described more fully below. The third group contains several varieties, one of which will receive a separate notice. Of the rest, it may suffice to say that with one exception they are ordinary types; this, however, though a well known one, has sufficient interest to warrant a slightly fuller description. The rock, which comes from the Hunza Valley, between Gulmet and Tashot,† is a dark lead-coloured schist containing garnets. The latter, on microscopic examination, are found to be a pale reddish colour, sometimes fairly regular in outer form, cracked, granular in structure, and often containing a fair amount of dusky enclosures. These have a somewhat dendritic grouping; the angles which the tufts make one with another are such as to suggest relations with the process * As has happened in many porphyroids (T. G. Bonney, Proc. Geol. Assoc.,' vol. 9, 1885, pp. 250-258).

Many fragments of this rock were lying along the bottom and slopes of the valley. Multitudes of garnets were found a little higher up the valley.-W. M. C.

of crystal building. Also tubes or fibrous cracks are present, arranged generally at right angles to the faces of the garnet. Rather irregular crystalline grains of yellowish staurolite with many enclosures occur; also irregular grains of magnetite, and numerous little patchy flakes of brown mica. These are all set in a crystalline matrix, consisting of white mica and (apparently) granular quartz, with usually a considerable quantity of opacite (probably graphite), a few small tourmalines (strongly dichroic, changing from a light to a brownish or dull greenish tint), and some small rutiles. A similar schist comes from near Askole, and one without garnets, but in other respects like these, from the south flank of Crystal Peak. The chief interest of these garnet-bearing mica schists is their very close resemblance to schists in the Lepontine Alps, as described by one of us, where the rock is a local variety of a dark micaceous schist, and it occurs, to his knowledge, at intervals for a distance of over 30 miles in a straight line along the chain.

*

With the mica schists we may mention, under the general name of sericite schists, several very much crushed rocks from Kamar nala, Mir, and the Dar Valley, Bagrot, and then pass on to a group of more or less calcareous schists, such as are developed in the Alps, and are there associated with quartz schists, green schists, and the aforesaid black garnet schists. In that chain they not unfrequently pass into crystalline limestones or dolomites, and rocks of this character also occur in Mr. Conway's collection. One or two contain malacolite, and some show distinct signs of having been affected by pressure.†

Passing on to the ordinary sedimentary rocks, we find a number of limestones, more or less impure, some containing fragments of other rocks, with schistose calcareous grits, besides argillites and slates, one or two of the latter resembling the slates of Llanberis (North Wales). A few of the specimens contain much crystalline material, so that it is difficult to decide whether they are very crushed dark schists, or slightly altered slates largely composed of detrital crystalline material. Rocks may be found in the Alps which present similar difficulties.

Sandstones, grits, and conglomerates occur; some of the gritty rocks show a cleavage, and certain near the Golden Throne probably contain volcanic materials. A conglomerate from Mapnun, in the Burzil Valley, contains a fragment of a quartz diorite, which obviously had been already modified by pressure when it was made into a pebble, indicating that in this mountain region, as in the Alps, earth move

* T. G. Bonney, 'Quart. Jl. Geol. Soc.,' vol. 49 (1893), p. 105, &c. We are informed by Mr. G. Barrow, F.G.S., that a similar schist occurs in the Central Highlands of Scotland.-T. G. B.

. T. G. Bonney, Geol. Mag.,' 1889, p. 483, and 1890, p. 536.

ments must have occurred long anterior to those which have produced the existing chain. Lastly, two partially altered sedimentary rocks are of some interest. The first, from a fallen fragment halfway between Samaiyar village and Strawberry Camp, on the left bank of the valley, is a blackish, compact, slightly-cleaved rock, not distinctly crystalline, in which are scattered several crystalline grains, the largest slightly more than one-eighth of an inch in diameter. This has one fairly well-marked cleavage, with a subvitreous, slightly oily lustre, and a second more imperfect, meeting it at an obtuse angle. The hardness seems to be slightly less than 5. Under the microscope the ground-mass is seen to consist of minute films of a sericitic mica mixed with a minute colourless mineral and granules of opacite and ferrite. In this are scattered larger irregular grains and plates of a black mineral, with raggedly outlined flakes of biotite, containing much of the ground-mass, some prisms (probably rutile), and two or three specimens of a larger mineral (probably the same species as that already mentioned). The best defined has two cleavages, one more strongly developed than the other, meeting at an angle of about 76°, and extinction takes place at an angle of 30°, or a little less, with the former. The crystals exhibit a rather irregularly outlined prismatic form, the sides being roughly parallel with these cleavages, and are crowded with minute materials, apparently identical with the ground-mass. This presents a slight resemblance to that of the ottrelite rock of the Forges de la Commune, Ardennes, and of one or two schistose rocks from the Alps, which do not belong to the most ancient group. Both the biotite and the abovenamed mineral appear to have been formed in situ at a time when molecular movements were not easy. We are unable to identify the latter with any mineral known to us, but it somewhat recalls to mind the "knoten and prismen" from certain Jurassic rocks in the Lepontine Alps, and even the couseranite from Vicdessos (Pyrenees). It does not seem to be tetragonal. We venture to suggest that it is a hydrous alumina-lime-silicate allied to the scapolite group. The matrix around the crystals is slightly coarser than elsewhere. Possibly the peculiarities in this rock may be the result of contact metamorphism. The other rock from near Trough Camp, on the right side of the névé, obviously contains rather angular fragments of white marble imbedded in a hard matrix, grey, speckled with dark green, in colour. The larger marble fragments are stained externally with limonite; many of the smaller are altogether brown. Microscopic examination shows these to consist alike of crystalline calcite, fairly coarse in the whiter parts, fine-grained in the iron-stained. Both structures are sometimes present in the same fragment, and their relations suggest that the fine-grained one comes from a * T. G. Bonney, 'Quart. Jl. Geol, Soc.,' 1890, vol. 46, pp. 213-221, 232–236.