134 MR. J. ROMANES ON THE [Feb. I9I2, 



detail in the pages dealing with that area (see pp. J 20-21). The 

 most important points in this connexion are (i) the bedding, very 

 clearly shown in the sands : in fact, in the finest deposits it is 

 lamination rather than stratification, and this bedding is apparent 

 even in the coarse conglomerates ; (ii) the extremely rounded and 

 waterworn condition of the boulders. There can be no doubt 

 that these are normal river-deposits, laid down prior to the out- 

 pouring of the lavas of the San Jose Valley. 



Passing on now to the deposits of the Rio Ueventazon, we find 

 that in general character they are similar to those of the Rio 

 Grande. They differ slightly, however, in being almost entirely 

 conglomeratic, without the thick interbedded sands which are so 

 conspicuous a feature in the section exposed at Cebadilla. The 

 deposits of the Rio Reventazon, as a rule, consist of very well- 

 rounded boulders up to several feet in diameter, embedded in a 

 sandy matrix and showing little stratification. 



At one place, however, in a railway-cutting about 2 miles above 

 La Florida, occurs a band of well-bedded sand about 18 inches 

 thick, intercalated with the conglomerate. This band strikes east 

 and west, and dips at an angle of 15° northwards. 



In places the rock-fragments are angular, and show no trace of 

 water-action. This is probably due to local land-slides having 

 added material to the river-deposits. The boulders are almost 

 exclusively composed of volcanic rocks. The Rio Reventazon has 

 cut deeply into these deposits, so that at La Florida a thickness of 

 about 200 feet is exposed above the present river-level. This 

 deposit forms a very well-marked, though fragmental, terrace 

 throughout a large part of the valley of the Reventazon. That 

 this terrace is formed of the ' Boulder Clay ' is well shown at La 

 F'lorida, although the actual deposits are obscured by vegetation. 

 The rise from the river is very abrupt ; but at a height of about 

 200 feet above the river a level bench is reached, with numerous 

 large well-rounded boulders scattered over it. Beyond this the 

 ground rises again, but no trace of waterworn stones is seen at 

 any higher level, so that the ' Boulder Clay ' obviously formed the 

 terrace. At one time the ' Boulder Clay ' must have completely 

 filled up the valley to this level, but recent river-erosion has laid 

 bare the old sides of the valley in many places. 



In no case did I see any evidence for post-Boulder-Clay dykes, 

 such as were described by Dr. Sapper.^ In some cases, however, at 

 first sight, igneous rocks seem to cut through the river- deposits ; 

 but closer investigation shows that these are simply buried spurs of 

 old rock exposed by recent erosion, leaving the ' Boulder Clay ' in 

 place on either side, and there is never the least trace of any 

 alteration of the ' Boulder Clay,' such as would have been occasioned 

 by igneous intrusions. 



The following appears to be the most probable sequence of 



^ Peterm. Mittb. Ergtiuzungsb. xxxii (1906j No. 151, p. 32. 



