Vii 
bricks, bits of pottery, and gravel, formed the general mass. This 
is by no means uncommon here, as it often happens that one floor 
has been formed on the top of another, and at varying distances 
apart. 
IN ow the season of 1852 was one of unusual wet, and Cirences- 
ter, like other places, was then inundated with water, and an 
examination of Roman sites in this place demonstrates that the 
antient floors, like too many of our modern ones, was just then 
below the water level, and even at Chesterton Terrace there was 
nearly a foot of water in the cellars, the floors of which are a little 
below the level of the lower Roman floor just adverted to; so that 
the upper floor would be entirely out of the way even of a much 
higher inundation than the present one, whilst a few inches more 
of water would have also covered the lower one. From these 
notes, therefore, we seem justified in drawing the following con- 
clusions :— 
Firstly—That Corinium, like Cirencester, suffered from occa- 
sional inundations, and that the occurrence of one floor above 
another in our Roman works, show how the antients remedied 
the inconvenience, 
Secondly—That these inundations were even deeper in antient 
than in modern times. This we shall have no difficulty in under- 
standing, if we consider the greater facilities these days offer for 
carrying off water by draining and other improvements. 
And, lastly, many of the appearances about our Roman sites 
seem to show that excess of water has been ofttimes repeated 
during the last fifteen hundred years, though at very distant and 
irregular periods. 
Nore.— The Notes upon which the above Paper was founded 
were principally made in the wet season of 1852. 
