Flames in Soiitli-cast Lake. 139 



injected with a network of molten lava filling the cracks not two inches from the sur- 

 face, and which, now plainly visible in the darkness, was a startling as well as a beauti- 

 ful sight. In daj'light the hot lava looks like black tar, and I have several times had 

 to pull my companions from the spot where they might be standing unconscious of the 

 silent black monster which was almost biting their feet, for it was nearly invisible on 

 the equallv black floor. 



In all of my previous visits the bank of the active pool had been at least twenty 

 feet above the lava surface, but now we were able to approach the southeast lake nearlj- 

 on a level, and the effect was much grander than usual. I have spent at various times 

 as manj' as ten nights in the crater on the banks of this and other similar lakes, and 

 have noticed blue and green flames playing over the cracks in the surface, biit these 

 seldom lasted longer than a few moments, and were not confined to any locality. Now, 

 on the contrary, on the top of a huge hummock which seemed to have broken from the 

 bank, was a cluster of blow-holes from which escaped constantly a large volume of gas 

 which burned with a bluish-green flame well shown in Mr. Furneaux's painting of the 

 lake made on this visit. (See frontispiece.) These jets were burning in the morn- 

 ing, and twelve hours after their volume was apparently unaltered. The pressure 

 evidently varied but slightly, and au}' increase in pressure did not seem to coiTcspond 

 to greater activitj- in the molten lava. With suitable apparatus it would have been 

 possible to have collected the gas before it was consumed. Its escape caused a noise 

 similar to that of a steamboat blowing off steam. The mention of steam leads me to 

 express a wish that those geologists who see in steam the prime cause of volcanic 

 action, could have been here, and have studied an eruption of the Hawaiian volcanoes. 

 A pailfull of water thrown into the southeast lake would have made more steam than 

 was present all the time we staj^ed in the crater. It is difficult to mistake a steamy 

 atmosphere for a very dr}- one; and then if steam was present in anj- quantity' in the 

 gaseous exhalations of Kilauea, the cold winds from Mauna Loa would soon precipi- 

 tate it as rain, when in fact this is the dryest part of the island. 



The ancient Halemauman or Everlasting House, where fires have been seen, 

 or whence vapors have escaped from time immemorial, was now replaced, I believe, by 

 the four lakes which occupj' the position of that single source. The guide and others 

 insisted that the northeastern of the lakes was the Halemaumau, and without renewing 

 ni}- survey, for which I did not have with me the necessary' instruments, I could not 

 positively declare that they were wrong, but I sighted from two of my monuments left 

 from 1S65, and comparing with mj- notes of that survey on my return to Boston, 



I found that the Halemaumau of that day occupied a position nearly .southwest of the 



[517] 



