.^94-] 265 fDodge. 
greatly diminish the amount of load furnished to that section of 
the river, so that the relatively greater carrying power acquired 
by the filtering of tlie stream by the lake would be ex|)ended in 
removing such alluvial deposits as might have been ]>reviously 
ilroiiptMl ill tlie lower reaches of tlie river. Such a snl)ii()inial 
barrier as is here described is formed by the juxtaposition of two 
alluvial cones in the Engadine and at least one place in India. 
i^cciDiiNTAL Alluvial Tekkaces. 
There might also be acci<lental variations during normal river 
development which in several instances would have an important 
inrtuence on the relations between the load and the carrying 
I)o\ver of a stream. The first of these may be dismissed with a 
few words, for the effect would be similar to that caused by a 
subnormal barrier in a stream, the difference being that in one 
case the barrier itself occurs as a subnormal feature of the river's 
life, while in the other case it is purely accidental. P''or instance, 
a timber dam, a beaver dam, or an artificial dam thrown across a 
stream might cause the building of an alluvial plain which would 
certaiidy be dissected and possibly terraced when that barrier was 
overcome. However, such a terrace would hardly be suHiciently 
long lived to cause the formation of well-marked topographical 
features. If the barrier were of glacial or volcanic origin, the 
effect might be felt for a longer time and the influence might be 
considerable, as we know for instance that many rivers in 
glaciated areas were seriously impeded in their course by the dams 
left during glacial times, and such dams have had an important 
modifying influence on the work of the streams of such areas. 
Lava obstructions occur in a few cases and specific instances of 
their effects ai-e therefore rare. 
Another glacial accident whereby an alluvial plain might be 
formed, is where a surcharged river suddenly emerges from 
beneath a glacier. The relief from the pressure of the over- 
lying ice sheet would cause an immediate deposition of the 
greater part of the waste carried while under pressure ; the 
removal of portions of this would form terraces. In this man- 
ner the terrace plain, if not the whole terrace, would be due to a 
glacial accident. An alluvial plain due to a much rarer glacial 
