Vol. 55.] GLACIAL PHENOMENA OF SPITSBERGEN. 685 



loose material by the advancing front of the Ivory Glacier were 

 borne out by phenomena of a similar character met with during 

 this expedition. Wherever suitable material existed, it was in- 

 variably picked up by the sole of the ice and sheared up gradually 

 through layer after layer till thrown out on the surface of the ice. 

 I would especially mention the debris so thrown out by the ice 

 on the south side of King's Glacier, and the englacial material 

 brought up against the nunatakkr mentioned above. 



lY. Drainage of the Ice-sheets. 



Where the ice-sheets are crevassed, as in the neighbourhood of 

 nunatakkr, no surface-drainage really exists, but everywhere above 

 the snow-line in summer, where the surface is unbroken, an 

 enormous quantity of water is constantly set free. This accumulates 

 to such an extent as to saturate the snow and form a spongy mass, 

 into which the traveller frequently sinks even when supported on 

 snow-shoes. This occasionally collects in some of the surface-eddies in 

 the ice and forms lakes, especially in the hollows formed on the lee 

 side of a nunatak. The resulting drainage gathers into rills and 

 water-channels, and eventually into voluminous streams which, in 

 King James's Land, have excavated channels several feet below the 

 surface of the ice. These channels, with their treacherous banks, 

 their slippery floors of hard blue ice, and their rushing torrents, 

 occasionally form serious obstacles to the exploration of the ice- 

 sheets. One of the interesting features to be noticed in connexion 

 with these streams is the abundance of rock-fragments occasionally 

 found in their channels ; these are derived from morainic material, 

 and during their passage downward become to some extent rounded 

 and waterworn before being deposited lower down the valley. 



In addition to surface-streams, englacial streams are also met with, 

 and it is in connexion with one of these latter that the most 

 suggestive deposits of this character occurred. About a day's march 

 from the coast, on the left bank of King's Glacier, we found a very 

 interesting example of an englacial stream, which originated in a 

 lake formed annually on the lee side of a large nunatak. During 

 the winter months the entrance to the channel was evidently 

 blocked by snow, so that the water, formed by melting snow in 

 spring, accumulated between the edge of the ice and the hollow 

 formed on the lee side of the nunatak. As the summer advanced 

 the snow-barrier melted away, and the water of the lake escaped 

 with great impetuosity down what may have originally been a 

 flaw in the ice, but which, when we saw it, was a tunnel 

 12 to 15 feet in diameter excavated in the solid ice. The direction 

 of the tunnel ran at right angles to the direction of the valley and 

 was plentifully strewn with moraine-material. This material was 

 derived partly from angular fragments which constantly fell from 

 the slopes of the nunatak, and partly from the englacial moraine 

 dragged up to the surface by the pressure against the nunatak. 

 The debris derived from both these sources had evidently been 



