longer diameter. Sometimes one finds the rock almost wholly made up of green uralite.

In places epidiorite takes on a schistose character, as in the outcrop near Spout Rollo, where it may be best examined in a small cutting on the road up the Glen. It also sometimes passes into hornblende-schist.

This latter is a rock of well-marked features. In it hornblende (unaltered) and plagioclase succeed one another in fairly well-defined layers, while red garnets appear as a secondary mineral throughout the rock. Its even fracture is strongly contrasted with the irregular break of epidiorite.

Even the non-schistose epidiorite is a true metamorphic rock, and is generally regarded as an altered coarsely crystalline basalt, but whether this basalt was originally the hornblende or the augite variety is a matter of doubt.

GLACIAL ACTION AND ITS RESULTS IN THE DISTRICT.

Mention has already been made of the numerous pebbles and boulders found on the surface or buried beneath it in all the river valleys and adjoining slopes throughout the district. These, on examination, are seen to be composed of the different kinds of rocks present in the Highlands-greywacké, granite, diorite, andesite, epidiorite, etc. In any given locality the relative number of boulders of a particular rock, such as diorite, or mica schist, will obviously depend on various facts. Such are—

1. The size of the areas from which these erratics have been drawn.

2. The nearness of those areas.

3. The durability of the kind of rock of which the erratics are

composed.

4. The capacity of the rock in its native place to resist disintegration by the different agents that act upon it.

Thus the most common erratic in the vicinity of Crieff is the quartzose greywacké, the source of which may be fairly presumed to be the extensive area of this rock, which lies not many miles distant. As we approach the low-lying parts of Strathearn erratics of diorite are seldom met with, since we are far from their native place; the out-crops of this rock are limited in extent, while it is a hard rock and resists disintegration. When detached, however, and carried away, it seems much more readily attacked by destructive agents than the fine-grained greywacké or granite, while natural acids more rapidly decompose its felspathic elements.

The degree of roundness of the erratics supplies us with valuable evidence as to their origin and history. Obviously the well-rounded pebbles and boulders of quartz usually occurring in conglomerates, or found loose on the sea-shore or in the beds of rivers, must, considering their hard and resisting substance, have been subjected to prolonged mechanical action since they left their native place amid the reefs of metamorphic rocks or the extrusions of granite. Their form is considered to be due chiefly to movement by water, either in the bed of a river or on a sea-shore. The vast majority of the erratics of Perthshire are, however, but imperfectly rounded. They do not need the mechanical action of shore waves to account for their present form. They suggest rather a rounding through weathering in conjunction with transit amid other loose material. Where the erratic is of volcanic origin its rounded form may be due to rotatory movement at the time of upheaval, or to subsequent action, or both; or, again, its form may be entirely due to that of the cavity in which it had origin, as in the case of agates and other nodules of igneous rocks. The rounding may be due almost entirely to a special kind of weathering, as in the spherulitic weathering of dolerite.

More or less regular grooving or striation is sometimes observed on the surface of erratics, as well as of masses in situ. Considerable care must be exercised in distinguishing between genuine striation through contact with harder substances, and the simulation of striation due to jointing and weathering. Instances of erratics exhibiting undoubted striation are seldom met with, but smoothed and grooved surfaces of rocks in situ are fairly common in the Lednock and Earn valleys.

Closely connected with the erratics, and with the smoothed and striated surfaces just referred to, are the numerous banks and mounds of gravel and sand observed in all the valleys over one hundred feet above sea-level. Of these, the area lying between the Allan, Knaik, and Machany, and the valleys of the Shaggie, Lednock, and Turret, afford the best examples. The mounds of these localities seem to be composed of materials of varying degrees of coarseness, from rough gravel, with intermixed pebbles and boulders, to fine-grained and stratified sand, with admixture of clay. Sometimes the clay appears in distinct layers, as in the railway cutting west of the Turret on the Crieff and Comrie line. Sometimes a calcareous cement has consolidated the layers of gravel and sand, and thus formed what we may call a "recent conglomerate." This latter may be well seen in a cliff on the right bank of the Shaggie, between Monzie village and falls.

In studying the causes of the various phenomena now mentioned,

it is instructive to compare with them the similar state of things observed in such countries as Norway, Switzerland, and Iceland, where perpetual snow clothes the higher mountains, and fills with its consolidated mass the upper parts of all the valleys leading from them. These masses, known as glaciers, are ever moving slowly downwards, melting completely at the lower end, and constantly being renewed at the upper by the fresh snows of each succeeding winter. These ice-floods are seen to be ever depositing the burden of rock débris they bear on their surface, carry embedded in them, or thrust forward beneath them as they move along. Should the climate of the country become warmer as the years pass, the glaciers retreat further up the valleys, and the morainic materials they once carried are left far behind. This is precisely what has happened in the valleys of France, Italy, and Norway, which radiate from the Alpine heights that are still covered with perpetual snow. In such places the glaciers, as the agents of transport, erosion, and striation, are observed actually at work, and no one who has visited these localities can doubt that it is the same cause which has carried the erratics, deposited the masses of gravel, sand, and stones, and produced the striation and rounding of surfaces that are observed in the valleys of Perthshire, although the glaciers themselves must have disappeared long ages ago. Travellers in the valleys of the Nile and the Euphrates find themselves surrounded by the remains of forty or fifty centuries. We who walk out in the smaller valleys of the Shaggie and the Earn, places unromantic indeed from the standpoint of the antiquarian, may see in the drift mounds and the erratics about us the ruins of a thousand centuries, while we may reflect that ages almost indefinitely more remote, perhaps a hundred thousand centuries, look down on us, so to speak, from the top of Ben Voirlich and the Aberuchill Hills.

Geologists are of opinion that the Grampian chain was once much more elevated than at present, and that its glaciers flowed outwards across the Ochil Hills into the valley of the Forth, while at a still more remote period the glaciers of the Pentlands, then mountains many thousand feet high, deposited their rock débris in lower Strathearn. Certain it is that the whole of Perthshire, except perhaps its highest peaks, has been subjected to the action of an ice-sheet that at one time spread over most of northern Europe. In this district the direction of ice movement seems to have been towards the south-east, as judged by the transport of erratics and the lines of striation.

Interesting in this connection is the distinction between valleys that have been formed or deepened by glacial action, and those which have more recently been carved out by water. Obviously

B