The Shingle Beach as a Plant Habitat. 89 
on the crest and scatter it over the back, thus maintaining the 
heighth of the bank. These two processes, the one anabolic, the 
other katabolic, are doubtless closely related ; and, taking the bank 
as a whole, their combined action must be fairly evenly distributed 
along its length, since the course of the bank from Abbotsbury to 
Portland follows a very even curve—though in detail, slight 
inequalities, not indicated in the higher scale maps, are discernible 
to the practiced eye. It is probable that these inequalities which 
exist on the seaward face would, if carefully correlated with the 
regions of activity and repose on the lee side of the bank, fully 
repay investigation. 
c 
Fig. 7. Profile of the Chesil Bank east of its point of separation from the 
land. The dotted line is to show the cutting back of a ravine and the transfer 
of the talus to the terrace. S., Storm Shelf; C., Crest; B., Back ; R., 
Ravine; T , Terrace; H.W.M.O.T., High Water Mark Ordinary Tides. The 
vertical scale is much exaggerated. 
To recapitulate*: so far as observation goes there can be little 
doubt that the Chesil Bank as a whole is creeping slowly shoreward, 
the onshore gales piling the shingle up, whilst, as a result of perco¬ 
lation, it makes its way down on to the terrace, which is continually 
advancing, now here, now there, across the Fleet, but on the whole 
with remarkable evenness. 
C.— UNDERC UTT1NG . 
Though of little moment in renewing the surface of the shingle, 
a third method merits passing reference, viz., the mobility which 
arises from undercutting on the lee side by tidal or other currents. 
In this way a talus is sometimes kept active for an indefinite period, 
as for instance on the Blakeney Bank at the point where the channel 
of the River Glaven (near Cley) impinges on the Bank. In some 
