332 HARRY 0. WOOD 



judgment was confirmed by Messrs. Waldron and Hardy, who 

 witnessed' its outrush from the high camp they occupied that night. 



Beginning gradually, probably before midnight, certainly as 

 early as i :oo a.m., there was noted a rapid decrease 1 in the brillancy 

 of illumination above the line of flow extending toward Honomalino. 

 By 3:00 a.m. this, as seen from our station, was quenched com- 

 pletely; there remained only the diffuse glow of the clouded sky 

 and the brilliantly lighted column of fumes rising at the fountain- 

 head, and a much subdued glow above the front of the flow. And 

 this last was decreasing rapidly. By 4:00 a.m. the earliest light of 

 dawn found the illumination at the fountainhead much like that of 

 the previous morning, with the lines of illumination above the 

 courses of flow almost wholly quenched. 



Shocks of earthquake continued to occur. Some were strong 

 enough to disarrange the struts and levers of the seismographs. 

 This made it inadvisable for the writer to be absent from the 

 observatory except at times when the director could be present; 

 so, after a brief visit to the front of the Honomalino branch, the 

 writer returned to Kilauea. 



During the evening and night of May 23-24 it was seen from 

 the observatory that the glow had extended far to the left of the 

 fountainhead in a direction estimated at south-southeast. No 

 such extension of the band of illumination had been seen in the 

 evening or night of May 22-23. This was due to the renewal of 

 flow toward Kahuku on a much larger scale than in the beginning. 

 Ultimately the Kahuku branches developed much greater magni- 

 tude than those in Kona. During this evening and night, 

 May 23-24, this glow was seen through shifting clouds, so that no 

 good opportunity for making a photographic record presented 

 itself in the earlier hours. And, owing to his complete loss of 

 sleep on the two previous nights, the writer undertook no prolonged 

 watch. During the evening and night of May 24-25 this illumina- 

 tion appeared elongated farther to the south, and perhaps slightly 

 abated in intensity. A photographic record of this (Plate III, d) 



1 It should be noted that moonrise occurred between midnight and one o'clock, 

 but that this decrease in illumination was positive, nevertheless, as evidenced by the 

 glow above the Kahuku tongue. 



