The influence which an irregular surface has in causing local deflections of an ice-flow is also well seen in Barra, where the striæ sometimes point some 5° or 10°, and sometimes 25° and even 35° north of west--these variations being entirely due to the configuration of the ground. This island is extremely bare in many places, more especially over all the region that slopes to the Minch. The Atlantic border is somewhat better covered with soil, as is the case with South Uist and the other islands already described. Vatersey, Saundry, Papey, Miuley, and Bearnarey, are all equally well glaciated; but as they show little or no low-ground with gentle slopes, they have preserved few traces of sub-glacial débris. In this respect they resemble the rockier and hillier parts of the large islands to the north. Till, however, is occasionally met with, as for example on the low shores of Vatersey Bay, and on the southern margin of Miuley. Doubtless, if it were carefully looked for it would be found sheltering in patches in many nooks and hollows, protected from the grind of the ice that advanced from the south-east. I saw it in several such places in the islet of Bearnarey, where the striæ indicated an ice-flow as usual towards the north-west. We have now seen that the whole of the Long Island has been ground, and rubbed, and scraped by land- or glacier-ice which has traversed the ground in a prevalent south-east and north-west direction. We have seen also that this ice attained so great a thickness that it was able to overflow all the hills up to a height of 1600 feet above the sea. It is needless to say that such a mass could not have been nurtured on the islands themselves. They have no gathering grounds of sufficient extent, and if they had, the ice would not have taken the peculiar direction it did. Instead of flowing across the islands it would have radiated outwards from the mountain-valleys. Where, then, did the ice come from?