THE SEINE, THE MEUSE, AND THE MOSELLE 
201 
while bearing these deductive criteria in mind. The Somme has 
latel}’’ been captured i)y the growth of a subsequent branch from 
near the elbow of the Soude; for, behold, at the little village of 
Ecury-le-Repos a sharp elbow in the course of this stream and 
a narrow trench for a moderate distance above and below the 
elbow. The Petit Morin is evidently the lower course of the 
Somme. On account of its diminished volume it is for the 
present unable to keep its valley clear of the detritus that is 
washed down from the steep valley sides in the upland, proba- 
bly near Boissy and Le Thoult; hence the great marsh of St. 
Gond and its extensive deposits of ]ieat about the head of the 
stream. The marshy head of the Petit Morin is still close to the 
elbow of ca})ture at Ecury-le-Uepos, and no obscquent stream is 
yet developed in this case. The change is clearly of recent date, 
• Look next at the Soude-Surmelin system. 1 fere the capture 
occurred long ago; there is no sign of a gorge at the elbow of 
capture. An obsequcnt stream, the Berle, about four miles in 
length, has grown toward the retreating escarpment of the u[)- 
land, and the head of the beheaded stream is now ten miles 
away from where it stood at the time when the(!aj)ture bad just 
taken j>lacc. Having lost its bead rather early in the history of 
