222 SCIENCE PROGRESS. 



especially the moraines of combe-glaciers, whilst Sty Head 

 Tarn is a good example of a lakelet due to the formation of 

 a dry delta, formed by a stream descending from the two 

 Gables. The freshness of the surface of one bank of the 

 stream, where much material has evidently fallen in, indi- 

 cates that the formation of this tarn may have been quite a 

 recent event, and it would be interesting to know whether 

 there is any proof of its non-existence in historic times. It 

 is evidently becoming filled up very rapidly by the deltas of 

 the streams which flow into it, and this also indicates a 

 recent origin. Although several of the lakeland tarns have 

 exits flowing over the barrier, the larger proportion of them 

 drain over solid rock, and naturally several of these occur 

 amongst the shallow valleys of the upland plateaux where 

 cols lower than the drift-barriers frequently existed. The- 

 tarn at Watendlath, if drift-dammed, requires the filling in 

 of a very deep caiion-like gorge, whose depth must have 

 been much greater than its width at the top; if the evidence 

 for the formation of the other tarns of Lakeland by drift- 

 dams were not so strong, one would feel inclined to admit 

 that Watendlath Tarn lay in a true rock-basin, but it is 

 hardly likely that this tarn had a special origin, and after 

 re-examining the surroundings I am persuaded that the 

 drift-filled gorge did exist. This would necessitate the 

 beheading of the old Lowdore- Watendlath Valley in pre- 

 glacial times by a ravine cutting back from Rosthwaite 

 which became filled with drift in the glacial period. It is 

 very desirable that some similar case in a less picturesque 

 area should be investigated by boring, for if narrow ravines 

 were common in pre-glacial times, a great many apparent 

 rock-basins may have narrow, sinuous, drift-filled gorges, 

 which might easily escape detection. The comparative 

 absence of narrow gorges in our upland regions is a some- 

 what surprising fact, and I am inclined to think that the 

 true explanation is that a large number of them, which 

 existed in preglacial times, have been filled up and com- 

 pletely concealed by drift accumulations. Under the 

 heading of lakes in depressions which have been blocked 

 by some material may be included those lakelets which lie 



