Deposits in the Magellan Territories. — Nordenskjold. 303 
not to find any deep-sea formation on the lowlands, analogous 
to the conglomerates on the shore.* 
This fact has no further bearing upon the question of 
whether the shingle formation is to be connected with a glacial 
period. On the other hand, the probability of that being the 
case has been considerably strengthened by the discovery I 
made in Terra del Fuego of immense masses of gravel of an 
exactly similar character existing in immediate proximity to 
the glacial boulder-clay described below. Moreover, in West- 
ern Patagonia, in the district watered by the upper reaches 
of the Coile river, I found intercalations in the shingle forma- 
tion of undoubted glacial origin. 
The supposition that commends itself most strongly to me 
is that the shingle is formed by big rivers whose sources were 
immense glaciers and which flowed through country with a 
comparatively level surface^ and hence, by reason also of 
the vast deposits, often altered their courses. This view of 
the case is the same as that propounded by von Haast as an 
explanation of the strata occurring in the Canterbury plains, 
New Zealand. t To appreciate the feasibility of this explana- 
tion, one must call to mind the fact that all data concerning 
the thickness of the gravel comes from the present-day river- 
valleys. It is only there that sections in the gravel are to be 
found. It is not even possible to establish whether this gravel 
occurs over the whole of the plain, since there it is often cov. 
ered over with later sedimentary deposits — "loess" — with in- 
tercalations of sand and gravel. If we now assume that the 
rivers of to-day have for the most part the same courses that 
their mighty predecessors in glacial ages followed, we should 
arrive at an explanation of the circumstance that the gravel 
exists to such a great depth in the walls of the present river 
valleys. 
*The statement made by Ameghino that intercalations have been 
met with in the bowlder clay, containing the shell of an Ostrea "of the 
same type as the Ostrea bourgeoisi," is very interesting, but is at the 
same time so incomplete that no conclusions can be drawn from it. 
Since the inquiries prosecuted by Hatcher it does not in any case 
seem likely that it can have any bearing on the question of the age 
of the formation. 
tj. von Haast. The Geology of Canterbury and Westland, 1879. 
jFurthermore also because if we assume two separate glacial periods, 
part of the gravel (to be) found in Patagonia probably belongs to the 
formations in the second period. 
