Pennine chain, but from a short spur which runs out eastward from the Portjen Grat. On the eastern side, the Val Vigezzo and perhaps the Isorno Valley are the only tributaries of any importance the mouth of the former being almost opposite to that of the Val Bognanco.

Under the circumstances, I thought that my best chance of running the antigorite to ground was to go up the Val Antigorio from Domo d'Ossola, keeping a sharp look-out for serpentine either as an erratic or in situ; and, if no sign of it appeared, to examine the pebbles in the bed, both of the Tosa and of its tributaries, since these

would show whether any important mass of that rock occurred in the Val Formazza or in the other valleys. The test, though obviously imperfect, is a fairly safe one, unless the outcrop be quite small.

In applying it, I had the kind assistance of my friend, the Rev. Edwin Hill, who about a fortnight before had joined me at Saas Grund, and thus had become as familiar with antigorite-serpentines as with those of the Lizard. A walk about Domo d'Ossola on the evening of our arrival gave some hope of success, for among the pebbles paving its streets serpentine of the Saasthal type is not

rare.

Next morning we drove up the Tosa Valley to San Rocco,

distant between 17 and 18 miles from Domo d'Ossola. The road, on leaving this town, runs over a stony plain, where also the pebbles gave promise of serpentine, but these were left unexamined for the present; and, after passing the junction with the Simplon road at Crevola, we entered the beautiful Val Antigorio. Here and there the road skirts knolls of a variety of gneiss, now commonly called after the valley: its chief constituents are quartz, white felspar, and rather abundant biotite; its cliffs and its mode of weathering suggest a granite, with a rather irregular jointing on a large scale, but the rock generally exhibits a fairlydefinite cleavage-foliation, and is thus extensively quarried, as in De Saussure's time, to make posts, steps, or flags for roofing, fencing, and paving. Gradually the sides of the valley become more craggy, and the alluvial plain by the Tosa narrows away on approaching the mouth of the Val Devero, the torrent from which rushes through a fine, though not very deep, gorge below Baceno, a townlet picturesquely perched on a headland between the two valleys. Beyond this, after passing through Premia, we enter the grander part of the Val Antigorio, with its magnificent crags, frequent waterfalls, and rich vegetation.2

At San Rocco,3 about half a league below Passo, the bed of the valley (perhaps a quarter of a mile wide from cliff to cliff) is fairly level, and thus boulders and pebble-beds are not unfrequent by the riverside. Here we began our examination, walking up the road for about 3 miles, nearly to Rivasco, examining the outcrops of rock and the huge fallen masses, now abundant by its side, and making our way at intervals (six or seven times in all, including once just below San Rocco) to promising places by the river. We saw nothing from the road but Antigorio gneiss, and it formed the great majority of the pebbles and boulders, a few being pressuremodified diorites, and a still smaller number a greenish granitoid rock. Not one was serpentine, from which I infer this rock to be either absent from, or very rare in, the Tosa Valley above Premia.

We had now to ascertain the quarter from which the lower part of that valley had received pebbles of serpentine. Early next morning we left San Rocco, and after crossing the Devero torrent came to a spot at the foot of a slope by the roadside, where, on the previous day, I had noticed a few blocks of serpentine (the last

1 That is to say, it is petrographically a gneiss, but is no doubt a pressuremodified granite. It is very fully described by Dr. S. Traverso in his Geologia dell' Ossola' 1895, pp. 59-69; also by Prof. H. Schardt, Archives des Sciences Physiques & Naturelles, ser. 3, vol. xxx (1893) p. 484.

The band of white marble shown in the Swiss map near Premia was not visible by the roadside; but I found the rock next day both on heaps of roadmetal and among the river-pebbles near Domo d'Ossola, nor was the Oberer gneiss with the micaceous variety containing unusually large garnets, of which I got specimens in 1860, at all well exposed (for a description see S. Traverso, op. cit. pp. 49-57). This gneiss differs considerably in aspect from the Antigorio gneiss, and reminded me of one variety common in the Saasthal. 3 Here, it may be mentioned, the Albergo Vesci affords very fair accommodation.

visible).' As at least a dozen were scattered over the few yards from which the surface-soil had been partly removed, and two or three of them were almost in contact, I infer that a fair number must be concealed by the vegetation. The blocks were angular or subangular in form, varying in diameter up to about half-a-yard, and the rock very closely resembled much of that in the Saasthal, being of a fairly dark-green colour, slightly mottled with blacker spots, so tough and hard that satisfactory specimens were difficult to obtain. It was but slightly schistose, developing under the hammer an irregular jointing. Under the microscope the rock is found to consist of antigorite, a residual augite, and an iron-oxide. The first needs little more than mention, as it resembles much of that described in our paper of 1905: it is practically colourless and non-pleochroic in a thin slice; it occurs in flakes, with hardly an approach to orientation and rather variable in size, the largest being about 025 inch in length, but most of them not exceeding the half of this. The augite, forming about one-eighth of the rock, occurs in irregular grains or granules, often with a slightly dusty' aspect, the former sometimes giving, for a space of about a fiftieth of an inch, fairly uniform polarization-tints, sometimes broken up, perhaps by pressure, into differently-coloured granules. The mineral is in process of conversion into antigorite, for it includes or is pierced by flakes of the latter, which bear no relation to the cleavage-planes in the few cases where these can be detected. The iron-oxide occurs in sporadic granules and irregularly-outlined grains, occasionally pierced by small flakes of antigorite. Examined by reflected light they have a general resemblance to magnetite, but the lustre seems to me not quite so bright as is usual; so possibly they are chromite or ilmenite.

Evidently these blocks are relics of the glacier which formerly descended the Val Devero-part of a scattered lateral moraine on its right bank. At the head of this valley the map records an outcrop of serpentine, measuring about 2500 yards from east to west, and 1500 yards from north to south, the apparent thickness of which is said to be several hundred feet'.2 This is crossed by the Geisspfad Pass (8365 feet) from Binn to Baceno. Some years ago my friend Mr. J. Eccles, F.G.S., kindly gave me two specimens of this rock. These consist mainly of an acicular green hornblende embedded in tale.

But as the former mineral

indicates pressure-metamorphism," and a schistose serpentine may be altered into a tale-schist, and as they were obtained near the junction with gneiss on the Swiss side, they are probably abnormal; for according to Dr. Preiswerk's account of the mass, to which I shall again refer, the bulk of it, which on the Italian side breaks up into separate sill-like intrusions, must often resemble the above-described

The place was nearly opposite the actual junction of the Devero torrent with the Tosa and at least 100 feet (I forgot to take a note) above it. John Ball, Central Alps' p. 254 (ed. 1866).

Quart. Journ. Geol. Soc. vol. xlix (1895) p. 94.

4 Geol. Mag. dec. iii, vol. vii (1890) p. 540.

1

rock. But the number of the boulders shows that there must be other outcrops in the Val Devero and Dr. S. Traverso states that serpentine occurs, among other localities, on the Pizzo del Cervandone, where it is said to contain both unaltered olivine and asbestos.

We had next to determine whether any quantity of serpentine was received from the Val Bognanco and the Val di Vedro, the streams from which, at their junction with the Tosa, are only about three-quarters of a mile apart, and thus, as their fans are practically continuous, a boulder brought down even by the former might very well be reckoned as occurring in the Val Antigorio. After returning to Domo d'Ossola, we examined the stony plain deposited by the torrent from the Val Bognanco, the main channel of which must, I think, have formerly flowed to the south of its present position. Here water-worn fragments of serpentine were abundant the larger (some 15 or 18 inches in diameter) more angular than the smaller, which were often flattened ellipsoids in shape; and the rock varying in structure from fairly massive to rather schistose. It was in fact indistinguishable from that so common in the upper part of the Saas Valley. Besides these, we found a few pebbles of a rock consisting of a fibrous green hornblende and rather saussuritic felspar, bearing some resemblance to one variety of the noted euphotide from the Allalin Glacier, and many of gneiss, which, however, generally was not of the Antigorio' type.

From this we went on to an extensive pebble-bed on the right bank of the Doveria torrent, and thus part of the fan from the Val di Vedro. Here serpentine was almost rare, and I saw only one pebble of the euphotide just mentioned, but the Antigorio gneiss was much more abundant, other varieties also occurring. The Swiss map shows two rather small outcrops of serpentine in the Val Bognanco, and none in the Val di Vedro; but I think that the facts stated above justify the conclusion that this rock must occupy a considerable area in the former, and may even have a few small outcrops in the latter valley; though we must not forget that the Tosa, when in flood, might contribute some pebbles to the stony plain which is usually the dumping-ground' of the Doveria.

After this we crossed the Tosa above Domo d'Ossola, where it is divided into three or four channels, and examined the pebbles on an island very near its grass-grown eastern bank. This, under ordinary circumstances, must be out of range of the two above-named streams, but some stones might have come from the Val Devero, or have been brought by the tributary from the Val Vigezzo (which we were unable to visit). Here pebbles of serpentine were not very common, but of the usual types; we also found some of a white or greyish marble, occasionally micaceous, representing that already mentioned; a few of the hornblendic euphotide; a fair amount of the Antigorio gneiss with some other varieties, and sundry

Geologia dell' Ossola' 1895, pp. 169-71.

1 2 Scale

100,000

In Dr. Traverso's map (scale) there are at least four, one being crossed by the torrent not far from the mouth of the valley.

crystalline rocks, including that called grüner schiefer by the Swiss geologists. I have examined under the microscope a specimen of serpentine from the Val Bognanco, and another from the shoal near the left bank of the Tosa, which might possibly have come from the Val Devero, but more probably from the Val Vigezzo, selecting the one to represent a moderately schistose variety, and the other one of the most schistose. The former rock is of a rather light greenish-grey colour, with small dark mottling. Under the microscope it is seen to consist of matted antigorite, the flakes seldom, if ever, exceeding 007 inch, and showing sometimes thornstructure,' sometimes a slight parallelism. The only trace of augite consists of cloudlets' of minute granules, seeming, with ordinary light, like a dust, only a very few of which show a fairly brightyellow tint with crossing nicols. The iron-oxide is much as is described above. The other specimen, rather conspicuously schistose, is of a greyish-green colour, slightly mottled with dark; under the microscope it is seen to consist of antigorite, with a little residual augite in aggregated granules, which often have a more flaky aspect and lower polarization-tints than is usual (perhaps from incipient decomposition), and with magnetite, both obviously crushed. The antigorite is small, seldom, if ever, exceeding ⚫004 inch, and shows a foliated structure, though not quite so conspicuously as I had expected. Enough to say that these serpentines of the Tosa valley, in both megascopic and microscopic aspect, so closely resemble those with which I was already familiar in the Vispthal, that I could not distinguish them, if without labels.

We see then that antigorite-serpentine, probably derived from an augite-olivine rock, is abundant in the region west of the Tosa from Foppiano to Domo d'Ossola, and is found in the Val Vigezzo,1 but its occurrence in situ in the Val Antigorio itself is very doubtful. Thus the original specimen may have been obtained, either from an erratic in the lower part of that valley, or from an actual outcrop in a tributary one (such as the Val Devero or the Val Bognanco), to which the name of the main valley has been rather inaccurately attached. The proof of its absence from the Val Antigorio is obviously not complete, but could only be made so by long and laborious work, for which I have now neither the strength nor the inclination, and I am content to leave the task to members of the Italian Geological Survey.

II. OTHER ANTIGORITE-SERPENTINES.

Since the publication in November 1905 of the paper by Miss Raisin and myself, I have obtained some more information on antigorite. About a year ago, Dr. J. M. Bell, Director of the New Zealand

1 Here, according to Dr. Traverso (Geologia dell' Ossola' 1895, pp. 166169), peridotites also occur.