THE SEINE, THE MEUSE, AND THE MOSELLE 
199 
arrangement to the actual arrangement? Undoubtedl}^ they 
would, and for the following reasons. 
The valley of the Marne being deeper than that of the Soude- 
Surmelin, the divide between the two would be pushed away 
from the larger and toward the smaller streams, and eventually 
the upper course of the Soude-Surmelin would be diverted by a 
growing side branch of the Marne (the lower part of the Sonde), 
and thus led to join that vigorous river, while the lower course 
of the Soude-Surmelin (the Surmelin) would remain as a dimin- 
ished, beheaded river. The side branch of the Marne, wliich 
causes the div'crsion, ]>elongs to the class of streams called sulm- 
qaent. Let us next look at the divide between the Soude-Surmelin 
and the Somme-Morin. At first, as these streams are of about 
equal volume, the divide between them would not be pushed 
significantly to one side or the other, but after the ca))ture of the 
Sonde by a branch of the Marne, the Soude would rapid I deepen 
its valley on the weak strata, and from that time forward the 
di videbetween the Soiideand theSomme-Morin would be system- 
atically pushed toward the latter. Eventually the upper waters 
of this stream would also be diverted to the Marne by the way 
