EFFECT OF STREAM WORK 419 



lation." This area was about 102 square kilometers (40 square miles), 

 and on this assumption the amount of ash deposited on this restricted 

 surface was about 510,000,000 cubic meters, or nearly one-eighth of a 

 cubic mile. Beckoning the catchment basin of the Wallibu at one-tenth 

 and that of the Kabaka at one-eighth of the area in question, we have 

 51,000,000 cubic meters as having been deposited in the former and 

 63,750,000 cubic meters in the latter. Guesses are hazardous, but it seems 

 probable that at least one-half of this large quantity has been washed 

 down from the slopes and carried into the sea. In 1903 I estimated 3 

 that not less than 5,500,000 cubic meters of ash had been washed out of 

 the Wallibu gorge alone in the ten months from the beginning of May, 

 1902, to the beginning of March, 1903. Much more has passed out by 

 the same route since. 



Manner of Stream Work 



The struggles of the streams with the debris began as soon as the erup- 

 tions deposited their loads. The rain from moderate showers sank at 

 once into the mass of the ash and produced no other effect at first than 

 to cause abundant explosions or secondary eruptions, as the water pene- 

 trated to the heated interior of the ash bed. From time to time these 

 secondary eruptions were of imposing magnitude, one observed by Mr 

 Curtis and myself on May 30, 1902, throwing its column of dust and 

 mud laden steam to an estimated altitude of about 1% kilometers. 4 I 

 can not, however, agree with the hypothetical section proposed by Mr 

 Curtis in the article to which reference is made or altogether with his 

 explanation of the phenomena. His section 5 is faulty in that the gorge 

 of the Wallibu does not have the V-shaped profile suggested therein, and 

 the arrangement of material in it was not that assumed by him. As is 

 indicated by figure 2, plate 43, the valley is broad in proportion to its 

 depth (that is, in its lower courses — the portion under consideration) 

 and is flat-bottomed, or nearly so. The present (1908) bottom is still 6 

 or more meters above the grade level reached before the eruptions; but 

 the walls are nearly or quite vertical, and the breadth of this part of the 

 valley is so great in proportion to the depth of ash still remaining in it 

 that it is evident that no V-shaped section can be present here. 



3 Comptes Rendus, IX Congres geologique international, 1904, p. 729. 



4 Hovey : Preliminary Report, etc. Bull. American Museum of Natural History, vol. 

 xvi, 1902, p. 343. 



Curtis : Secondary phenomena of the West Indian volcanic eruptions. Journal of 

 Geology, vol. xi, February-March, 1903, p. 200. 



6 Loc. cit., p. 211. 



