WATER AND VOLCANIC ACTIVITY DAY AND SHEPHERD. 303 



be regarded as conclusive, this lava is not appreciably affected by 

 water at the temperatures which obtain in the lava lake up to the 

 time when the water leaves the lava and is discharged into the air. 

 Pending the acquisition of more detailed data, therefore, we may 

 leave this question in abeyance, with reasonable confidence that it 

 will be found to be in full accord with the fact otherwise establislied, 

 that water is present and participates actively in the volcanic activity 

 at Kilauea. 



ORIGIN OF THE WATER. 



If we now grant that water is present as an active ingredient of the 

 liquid lava, in view of the positive character of the evidence offered, 

 then it becomes pertinent to inquire whether this water is of mag- 

 matic or of meteoric origin. Obviously, to this question no such 

 positive answer can be returned as that which was offered in support 

 of the main thesi's of this paper. It is conceivable (1) that water 

 may have entered by infiltration of sea water from the surrounding 

 ocean, or (2) through more or less deep-seated infiltration of water 

 originally meteoric, or, finally, (3) that it may be considered strictly 

 magmatic in character and an original constituent of the lava. 



The volcanoes of Hawaii are completely surrounded at no great 

 distance by the sea, Mdiich rises on their flanks to a height of 15,000 

 or 16,000 feet, according to charted soundings and the observations 

 of Dutton.^ The crater of Kilauea is about 15 miles from the nearest 

 approach of sea water, as recorded by the most modern surveys. The 

 rock is for the most part porous in high degree. Above sea level rain 

 falls almost daily on the island up to elevations of 7,000 or 8,000 feet. 

 Most of this meteoric water is deposited on the windw^ard side^ of 

 the mountains and the leeAvard portions are desert or nearly so. The 

 Kilauea Crater is situated on the flank of Mauna Loa at an elevation 

 of about 4,000 feet above the sea and is exactly on the ridge Avhich 

 separates the region of rainfall from the desert of Kau. It is some- 

 what misleading to assume with Dana that the rainfall at the crater 

 is comparable with the rainfall at Hilo, the nearest considerable 

 town where meteorologic observations are made. Hilo is to wind- 

 ward of the crater and at sea level. At the Volcano House, still 

 some 3 miles to windward of Halemaumau, the rainfall tables lately 

 published by the United States Geological Survey give the aimual 

 average for the years 1909-1911 as 78.7 inches at the Volcano House 

 and 136.5 inches at Hilo. It is also true, though it can not yet be 



1 C. E. Dutton : Hawaiian volcanoes. Fourth annual report United States Geological 

 Survey, 1882-83. 



= It will, of course, be recalled that the islands of the Hawaiian group are within the 

 trade-wind belt, and that the direction of the wind is very nearly constant throughout 

 most of the year. 



