340 L. A. Thurston — Recent eruption of ' Kilauea. 



leaving a clean line of fracture ; the line of this area, marked 

 B, B, B, was continuously leaning over and falling into the 

 lake. From about 12 until 8 in the evening there was scarcely 

 a moment when the crash of the falling banks was not going 

 on. As the level of the lake sank, the falling rocks of the 

 banks, undermined by the escape of the lava, caused a constantly 

 increasing commotion in the lake as they struck the surface of 

 the molten lava in their fall. A number of times a section of 

 the bank from 200 to 500 feet long, 150 to 200 feet high, and 

 20 to 30 feet thick would split off from the adjoining rocks, and 

 with a tremendous roar, amid a blinding cloud of steam, smoke 

 and dust, fall with an appalling down plunge into the boiling 

 lake, causing great waves and breakers of fire to dash into the 

 air, and a mighty " ground swell " to sweep across the lake 

 dashing against the opposite cliffs like storm waves upon a 

 lee shore. 



Most of the falling rocks were immediately swallowed up by 

 the lake, but when one of the great downfalls referred to occur- 

 red, it would not immediately sink, but would float off across 

 the lake, a great floating island of rock. At about 3 o'clock 

 an island of this character was formed, estimated to be about 

 125 feet long, 25 feet wide, and rising 10 to 15 feet above the 

 surface of the lake. Shortly after, another great fall took 

 place, the rock plunging out of sight beneath the fiery waves. 

 Within a few moments, however, a portion of it, approximately 

 30 feet in diameter, rose up to an elevation of from 5 to 10 

 feet above the surface of the lake, the molten lava streaming 

 off of its surface, quickly cooling and looking like a great rose- 

 colored robe, changing to black. These two islands, in the 

 course of an hour, floated out to the center and then to the 

 opposite bank. At 8 in the evening they had changed their 

 appearance but slightly. They had disappeared the next 

 morning. 



About noon the sinking lava disclosed the fact that the small 

 extension at the right of the lake was only about 80 feet deep, 

 and it was soon left high and dry, simply a great shelf in the 

 bank, high up above the surface of the lake. As the lava sub- 

 sided, most of the surrounding banks were seen to be slightly 

 overhanging, and as the lateral support of the molten lava was 

 withdrawn, great slices of the overhanging banks on all sides of 

 of the lake would suddenly split off and fall into the lake beneath. 

 As these changes took place the exposed surface, sometimes 100 

 feet across and upward, would be left red hot, the break, evi- 

 dently having taken place on the line of a heat- crack which 

 had extended down into the lake. 



About 6 o'clock the falling bank adjacent to the hill worked 

 back into a territory which, below 50 feet from the surface, 



