Vol. 68.] GEOLOGY OF COSTA RICA. 133 



of alluvial flat which stretches a long distance inland. The 

 coastal topography in this area corresponds, to a really remarkable 

 extent, with the description given by Mr. Hayes ^ of the coast- 

 line farther north-west in Nicaragua. He describes the same 

 phase of submergence followed by much ' alluviation,' which has 

 considerably modified the original outlines. 



VI. The ' Boulder Clays ' of Costa Rica. 



Of the surface -deposits existing in the country, the most 

 interesting are certainly the so-called ' Eoulder Clays/ which occur 

 often to a very considerable depth, both on the Atlantic and on the 

 Pacific slopes. 



The occurrence of deposits similar to the Eoulder Clays of 

 Great Britain in Nicaragua was described in considerable detail by 

 Thomas Belt.'^ He unhesitatingly ascribed to these a glacial origin. 

 Similar deposits iu Costa liica are recorded by Mr. 11. T. Hill,^ who 

 puts i'orward a theory that all these deposits have been formed by 

 the rolling of material from the steep slopes of the volcanoes down 

 into the valley. The large fragments of rock are supposed to 

 become rounded in the process, while the finer material is weathered 

 and forms the matrix. Having had a good opportunity of seeing 

 these deposits both on the xltlantic and on the Pacific slopes, I find 

 it impossible to agree with either of the theories mentioned above. 

 As regards the first, nowhere did I see the least trace of glacial 

 action, nor anything t,o suggest it, beyond the similarity of the 

 deposits to Boulder Clay. The second theory, that of E. T. Hill, 

 quite fails to explain the observed facts, as regards both the 

 character of the matrix and boulders and the mode of occurrence of 

 the deposits as a whole. His theory may partly explain the nature 

 of the agents in certain local examples ; in those cases, however, on 

 which he lays great stress, not only have 1 been unable to find any 

 evidence in favour of his theory, but there is direct and indisputable 

 evidence against it. 



So far as my experience goes, the formation of these Boulder 

 Clays cannot be ascribed to one cause alone, but several processes 

 must be taken into account. These may be classified under three 

 headings, in order of their relative importance : — 



(1) Torrential deposits ; 



(2) LandsUdes and scree-deposits ; 



(3) Spheroidal weathering and rearrangement by percolating water. 



Taking the most important first, namely, torrential action, 

 deposits thus formed are best seen on the Pacific side in the gorge 

 of the Rio Grande at Cebadilla, and on the Atlantic side along the 

 valley of the Rio Reventazon from Paraiso to the coastal plain. 

 Those of the Rio Grande gorge have already been described in full 



^ Bull. Geol. Soc. Amer. vol. x (1899) pp. 289-93, 339-40. 

 ^ ' The Naturalist in Nicaragua ' London, 1874, pp. 259-64. 

 3 Bull. Mus. Comp. Zool. Harvard, vol. xxvni (1898) p. 223. 



