422 E. O. HOVEY WALLIBU AND RABAK1A GORGES 



streams. Our explanation was that the overloading of the streams with 

 debris caused temporary periodic damming or checking of the water, the 

 waves being due to the accumulated water overcoming the obstruction. 



In June, 1908, the upper level of the floodplain of the Wallibu at the 

 point where the river leaves the old lines of bluffs was about 5 meters 

 above the level of the floodplain before the recent series of eruptions 

 began, as was shown by its relation to a bit of old stone wall that, accord- 

 ing to my guide, an intelligent black man, formed a part of the water- 

 works pertaining to the Wallibu sugar estate which was ruined by the 

 eruption of 1812. The bluffs were about 215 meters apart at their bases, 

 as determined by pacing. The vertical section of the new ash at this 

 point, therefore, may be taken at 5 meters by 215 meters. About 2% 

 kilometers from the sea the river in June, 1908, just as in March, 1903, 

 was flowing in its old channel in the decomposed ash. At this point the 

 bed of the gorge is about 50 meters across. 



The evidence of the successive stages in the reexcavation of the gorge 

 is given in the terraces bordering the stream at several points. The best 

 series of these, perhaps, is at a distance of about 2 kilometers from the 

 sea (see figure 2, plate 43), where, in June, 1908, six terraces recorded 

 the height to which each successive dry and moderately rainy season filled 

 the gorge, and indicate the down-cutting due to each succeeding season 

 of torrential rains, the "dry" seasons being periods of aggrading stream 

 action and the rainy seasons being periods of degradation. The highest ter- 

 races probably indicate the level of ash filling due directly to the eruption 

 clouds. The slope of floodplain and terraces near the sea was measured 

 at 3, 3.5, and 4 degrees in the Wallibu and other gorges. The condition 

 of affairs between the bluffs at the mouth of the Wallibu, when Doctor 

 Anderson was there in the spring of 1907, 7 was completely altered by 

 June, 1908. Doctor Anderson's beautiful photograph shows a broad ter- 

 race 9 meters high of water-sorted ash at the right (north) side of the 

 gorge, and extending seaward beyond the bluff line, younger terraces ap- 

 pearing at lower levels on both sides of the stream. The stream was 

 flowing along or near the left wall of the gorge. In June, 1908, there 

 was not a trace of these terraces to be seen, all the material having been 

 swept away, and when the stream flowed on the surface it occupied a chan- 

 nel at the immediate base of the great bluff forming the right wall of the 

 gorge. Beginning just within the gateway formed by the bluffs and ex- 

 tending seaward for 200 meters more or less, there was, in June, 1908, a 



7 Philosophical Transactions of the Royal Society of London, series A, vol. 208, p. 

 278, pi. 11. 



