512 S. POWEHS TECTONIC LINES IN THE HAWAIIAN ISLANDS 



from the east end of the central portion southward; and the Kipahnhi 

 rift, concealed b}^ a jungle, stretches in an easterly direction from near 

 the east end of the central portion and is really a continuation of that 

 portion, but is completely separated from it by a comparatively narrow 

 wall which effectually prevented exploration and even discovery for a 

 number of years. The Koolau and Kaupo gaps, and especially the latter, 

 are flooded by recent flows. The few men who have succeeded in going 

 from the Koolau Gap into the Keaenae Valley to the sea report that the 

 valley is composed of a series of amphitheatral depressions between the 

 top of the mountain and the Ditch Trail, where there is a similar amphi- 

 theater 3 miles long and 1 mile wide. Down the head of this amphi- 

 theater waterfalls drop from heights of from 500 to 1,000 feet to the flat 

 valley floor. Descrij)tions of similar structural features of the Kipahulu 

 rift near its head are given, but near sealevel at Kipahulu there is no 

 large valley (due perhaps to concealment by later flows), and one would 

 never suspect the existence of the rifts far up the mountain slope. 



Cross^^ argues that the Koolau and Kaupo gaps were formed by valley 

 erosion rather than by block-faulting. If so, why does the Keanae A^alley 

 (Koolau Gap) grow smaller downstream when it heads in a, trench so 

 dry that no vegetation grows? The Kipahulu rifts are pronounced, ac- 

 cording to authentic reports, near the Haleakala rent, and certainly the 

 walls of the Kaupo Gap do not represent products of erosion. 



Dana suggested that a triangular portion of the volcano had split out- 

 ward or slightly subsided. If the former were the case, the corner of 

 the triangular block opposite White Hill would have originally fit into 

 White Hill, the summit of the mountain, and the sides of the Kaupo 

 and Koolau gaps should roughly fit; but, at present, they would not so 

 fit (even allowing for some faulting) if placed in the correct positions. 

 In case the triangular block had subsided, the corner opposite White 

 Hill should be appreciably lower than that hill (it is slightly lower), and 

 the shoreline should not show outward displacement around the eastern 

 side of the island. This shore is slightly protruding. 



Another possibility is here offered: that the rents and the gaps Avere 

 formed simply as graben similar to the ones which appear to have in 

 large part made the magnificent Keanae and adjacent Honomanu valleys. 

 In the heads of neither of these valleys is there any visible evidence of 

 any splitting of the mountain as suggested by Dana. Somewhat in the 

 manner in which the summit sink of Kilauea and Mokuaweoweo were 

 formed, the Haleakala rent may have originated along perpendicular lines 

 of weakness, while the portion of the mountain between the main trough 



27 U. S. Geol. Survey, Prof. Paper No. 88, 1915, pp. 92-93. 



