98 



Kilauea and Manna Loa. 



supply except from the rainfall on the sloping sides of the crater; it is about twent}' feet 

 deep, and although viewed from above the water looks very green, it is quite sweet and 

 colorless in a glass. The natives assured me that its waters become j'ellow and then black 

 during an eruption of Kilauea, a statement I much doubt. An adjoining crater contains 

 many coconut trees: the walls between the craters of this group are thin as some of 

 those in Hualalai, and like those pit craters are walled with stony lava and not tufa. 



Half a mile from this group is a cone about two hundred and fifty feet high, 

 and crowned with an ancient heiau or temple, and a clump of coconut trees. This 

 cone is largeh' composed of lava, and is doubtless of great age as the soil is several 



Fig. 65. VIEW OF THE PUNA CRATERS FROM THE XORTH. 



feet deep upon it in some places. At its base is a large cleft in the rock some three 

 hundred feet long and sixty wide, in which is a remarkably clear pool of warm water, 

 twenty or thirty feet deep and of a temperature at the time of my first visit of 90°. 

 Thirt3^-four years after I found it five degrees above the air temperature. At a later 

 visit to test more accurately I found the pool occupied by a gang of Japanese field 

 hands. The water is not mineral to the taste, but the dark-colored bodies of natives 

 swimming in it seem almost wax-colored, and a white man in the pool resembles marble. 

 The sound of water trickling down within the cliff is distinctly audible after a rain. 

 Three-quarters of a mile from this is a deep narrow cavern into which one may climb 

 guided by the natives with their bambu torches. There is a steep descent of nearly 

 fifty feet to a pool of very warm water which is said to extend more than half a mile 

 under ground.'" All along the shore for twenty miles, warm springs are common near 



"In 1864 the writer saw several natives swim nearly that distance holding their torches clear of the steaming 

 water, and the indistinct view thus obtained of the cavern, led to the conclusion that it was a deep crack, over which 

 a subsequent lava-flow had formed a roof. [4761 



