Terraces and River Valleys— Spencer. 155 
the Erne (a distance of about 80 land miles) and the Shan- 
non river (for 100 miles) to the amphitheatres indenting 
the edge of the platform. So also the course of the sinuovis 
channel from the Irish sea is given for a distance of 300 
miles before it falls over the edge of the shelf into its can- 
yon. It is a relatively shallow feature from 50 to 240 feet 
in depth. The course of the old river of the English chan- 
nel is more interesting. Prof. Hull shows its course from 
near the straits of Dover for nearly 400 miles to "the edge 
of the platform. For 70 miles of its course it has been 
known as the Hurd deep," having a width of four to five 
miles, and greatest depth of 354 feet below the general 
level of the floor of the sea (further submerged by 216 
feet). Below the deeper section the channel is less ap- 
parent, probably owing to the silting up by sediments. 
Hull considers that the Hurd deep portion has been kept 
open by tidal currents, as this is the narrowest section of 
the English channel north of cape de la Hague. It is cut 
down into solid rock and is bounded by precipitous cliffs. 
There are tributaries from France and England. The 
lower portion of the channel is again open. It abruptly 
merges in a canyon, with walls of rock 4,000 feet in hight, 
cut into the great continental declivity and shelf on which 
it is submerged to the depth of 100 fathoms. An adjacent 
and larger embayment to the east ma}^ have been the former 
course of the channel. After passing the edge of the shelf 
Prof. Hull has used only isobaths of 250 fathoms. Had he 
put them at 250 feet apart more detail would have been 
obtained. The British rivers are often flowing over refilled 
channels, reaching much below sea level. Prof. Hull con- 
sidered the British platform to have been planed down by 
wave action, and subsequently depressed. He compares 
the features with the "drowned plains," escarpments and 
river valleys lying outside of the North American coast. 
He points out the absence of the lower terrace such as that 
of the Blake plateau of the American coast. Prof. Hull at- 
tributes the fashioning of '^'the escarpments as mainly due 
to wave action undermining the cliff during prolonged 
pauses in the process of elevation or subsequent depres- 
sion." The age of the platform is assigned to the Mio- 
