THE TRANSVERSE VALLEYS. 225 



the San Julian no stream at present ever flows, its waters having been long 

 since captured by the Rio Chico of the Santa Cruz. 



These valleys had their origin before the last submergence of Patagonia 

 and are remnants of former drainage systems that existed prior to that 

 submergence and to which the present drainage systems of Patagonia are 

 largely though not entirely conformable. 



My reasons for assigning the origin of these valleys to a period earlier 

 than the present are : (i) The fact that the terraced slopes which termi- 

 nate the series of benches encountered at successive altitudes, as one pro- 

 ceeds inland from the coast, are continued also along the sides of all the 

 great transverse valleys. In many places these terraces exhibit all the 

 characters peculiar to a beach formation and leave little doubt that the 

 valleys, on the opposite sides of which they are now found, were, at the 

 time the terraces were forming, occupied by great inlets from the Atlantic, 

 (a) In several instances during my travels in Patagonia, I observed that 

 lava flows which had been erupted before the deposition of the Shingle 

 formation, while having had their source in some volcanic cone or dike at 

 the summit of the high pampa, had flowed out, not only to the extreme 

 limits of the higher pampa, but down over the slopes and into the valleys. 

 Remnants of such lava flows are to be seen at several localities, both in 

 the valleys of the Santa Cruz and of the Rio Chico north of that stream. 

 The conformation of such a lava flow, alike to the surface of the high 

 pampa, to the slopes and, to a certain level at least, to the sides of the 

 valley, is conclusive evidence that, before the origin of the lava stream, 

 the valley had already been cut to a depth at least equalling that of the 

 lower limits of the lava. That the origin of the lava of several such 

 streams antedated the Shingle formation and the underlying Cape Fair- 

 weather beds, was also conclusively demonstrated by the fact that, up to a 

 certain level, which by the way is a correct register of the maximum 

 amount of the last submergence, the basalts were overlaid by deposits 

 belonging to the Cape Fairweather and Shingle formations. An excellent 

 example of this may be seen in a lava flow on the north side of the Rio 

 Chico of the Santa Cruz, some fifteen miles above Sierra Oveja. At this 

 point the tableland to the north of the river rises to an altitude of some 

 fourteen hundred feet above the bottom of the valley, and its summit 

 and slopes, to within perhaps two hundred feet of the bottom, are 

 covered with a continuous sheet of lava; this, in turn, is covered with 



