90 JO^'ES : explorations' of a cave at elbolton, near thorpe. 
miners said the floor had been "well puddled over" on the way to 
the further recesses of the cave. In the shaft of the cave I think 
there is also evidence of footholds for the dwellers to climb in and 
out of the hole. 
I am, dear Sir, yours faithfully, 
To James W. Davis. Edw^ard Jones. 
OS A LAKE-DW^ELLING RECENTLY DISCOVERED AT PRESTON (lETTER). 
by rev. e. maule cole, m.a. 
Dear Sir, 
A few years ago Mr. T. Boynton discovered a lake-dwelling at 
Ulrome, in Holderness, the first of the kind found in England. I 
have just had an opportunity of exploring the works, which have been 
for some time in progress, for diverting the course of the Ribble at 
Preston, in Lancashire, and constructing large docks in the marshes 
which border on the river course. 
The excavations are at present at a standstill, and will probably 
remain so for a considerable time. The sections exposed, 30 feet 
deep, are extremely interesting. Every trace of boulder clay, which 
covers the ¥j\de elsewhere to a great depth, was ages ago removed by 
the river, and on the bared new red sandstone w'ere deposited thick 
beds of gravel and sand, full of estuarine shells. In these gravels, at 
various depths, lie scores of huge blackened trunks of trees, with 
their ends mostly in the same direction — downstream. There is no 
trace of soil ; they have all, without exception, been drifted into 
their present position, a position which does not appear to have been 
altered by any change in the river's course since the gravels were laid 
down. 
Whilst tracing the line of trees I caught sight of some piles 
standing vertically in the gravel beds, and on closer examination 
found that they had been driven in to a depth of from 8 feet to 15 
feet. The ends were roughly pointed. For a depth of 2 feet or 3 
feet at the top a quantity of brushwood had been laid horizontally 
between the piles, so as to make a solid and firm floor. The length 
of the floor was about 17 yards, but I could not ascertain the width 
