Terraces and River Valleys — Spencer. 163 
and would imply tremendous changes all over the earth 
since middle Tertiary times. These changes in the moun- 
tains have been more or less studied, but their counterparts 
beneath the sea are less accessible and have hitherto been 
largely passed over. Now they have their innings, and 
Prof. Hull has been a creator of a new chapter of earth's 
science. Hull it was who specially surveyed the Jordan- 
Arabah valley, which is the continuation of Prof. J. W. 
Gregory's "great rift valley." This circumstance specially 
fitted him for judgment on the submarine valleys, although 
the evidence was less complete than he would wish for. 
Yet in the face of Prof. Hull's evidence Prof. Gregory made 
the following criticism : 
"Prof. Hull's theory has the attractiveness of simplicity, but 
it involves the very improbable assumption of a 9000 feet elevation 
of the coast, whereas there is no proof of such elevation on the 
adjoining shores. This improbability renders necessary a careful 
consideration of the alternative theories of the formation of these 
submerged canyons. There is no doubt that some of these chan- 
nels are submerged river channels, but in many cases the explan- 
ation is doubtful. Thus Marcel Bertrand has explained the 'Hurd 
deep' as a line of subsidence by folding, and that suggestion must 
be refuted before we can accept the Hurd deep as a river channel. 
Another theory explains these channels as canyons of deposition 
instead of canyons of erosion. For instance there is the canyon 
off the mouth of the Congo, which Buchanan attributes to materials 
brought down by the river being deposited on either side of the 
mouth. In some cases these canyons occur where it is quite im- 
possible that they could have been formed by subsidence; e. g. at 
the eastern end of lake Geneva, where the Rhine flows into the 
lake, there is a canyon which cannot have been formed by erosion. 
Are Prof. Hull's canyons genuine canyons, or are they to be ex- 
plained by other theories? In the case of the 'Hurd deep', there is 
the evidence that it is a probable line of warping. In regard to the 
Irish channel river there is no single depression such as Prof. 
Hull's diagrams suggest, but a series of elongated banks formed 
by the action of currents * * * The paper would have been 
more convincing had it given a monographic treatment of one or 
two cases, instead of a general survey of a wide field. The exist- 
ence of some submerged river channels is probable, but as there are 
alternative theories explaining the facts, every case has to be 
judged on its own merits." 
To this, Prof Hull replies that "questions of probabil- 
ity or otherwise in natural science subjects are matters 
which T (Hull) cannot possibly recognize." The alterna- 
ti\-e theory of the Congo based on the building up of the 
continental shelf and slope to a hight of 6,000-7,000 feet 
with the channel of the river, doing this work kept open is 
sim])l\' an im])ossibilitv, for the power of the current soon 
