OF THE HAWAIIAN ISLANDS. 461 



is not brought to the surface by the rise and expansion of inflating vapors. That vapors 

 accompany it is perfectly true, but the molten mass is often drawn from a cone some yards 

 below the surface of the liquid which emits gases, and the product is smooth and glassy ; not 

 a bubble disturbs its rapidly hardening surface, and often no vapor accompanies its egress. 



That the lava is not much above the melting point on the surface of the Halemaumau 

 and other pools in Kilauea, is evident from the great rapidity with which it becomes gran- 

 ular. Where the substance is drawn out by gravity the granules are very distinct, even 

 when the lava was liquid enough to form pendants a foot long. Several stalactites formed 

 from melted lava exhibit this structure admirably. The lava is melted as the Rowley rag is 

 melted in the furnaces of the Messrs. Chance, and is not a simple solution in some vehicle. 

 Dolomieu, in speaking of volcanic fire, says : "II produit la fluidite par une espece de dissolu- 

 tion, par une simple dilatation qui permet aux parties de glisser les unes sur les autres, et 

 peuketre encore par le concours d'une autre matiere qui serf de vehicule d la fluidite." ' Else- 

 where he supposes this vehicle to be sulphur, a supposition not more improbable, judging 

 from Kilauea alone, than the theory of Mr. Scrope, that water is the interstitial fluid which 

 imparts mobility. 2 The fusion is perfect as seen in the Pele's hair, and when the lavas 

 granulate they do so without any disengagement of vapor. I have seen the streams or rills 

 of lava moving with such entire freedom from any thing like smoke, that had I not been 

 watching, they might have passed near me unnoticed. Wherever the melted rock passes over 

 combustible matter, or through swamps, the vapor generated is sufficient to convert the 

 surface and mass also into a porous rock. It must not be inferred that gases never inflate the 

 lava in the crater. Much of the surface overflow is, as has been said, a spongy scum, but 

 it is only the surface overflow that is so porous, and if water is necessary — if vapors are 

 necessary to the elevation of the matter — these gases must be in a state of greater tension 

 lower down in the column, and ought to froth out when an outlet is provided ; and we should 

 have, with the pressure of a column of lava three thousand feet in height, a result similar 

 to that produced by the sudden expansion of liquid carbonic acid, and a black or green 

 snow would surround the vent. Something like this seems to occur, but in quantities 

 wholly inadequate to the agency supposed. Around the orifice of 1859 on 

 Mauna Loa are several cartloads of a very light porous substance, called by the 

 natives limu or moss. It is dark green and smooth on the surface of the irregularly rounded 

 nodules, while the interior is light green. The same limu forms large beds on the leeward 

 bank of Kilauea, and there its origin is explained ; it is simply the scum from the Halemau- 

 mau blown away by the wind, not by liberated gases, and its formation is going on with that 

 of Pele's hair. 



The structure of the fresh lava is not easy to understand, some parts of the same stream 

 haviner a clear rin°; when struck, others being dead and flat. The mode of granulation 

 certainly has much to do with the phonetic qualities of lava, as the a-a or simple 

 granules sound quite differently in different places, although of the same composi- 

 tion and appearance. The a-a of Kilauea contains but little olivine, and that in small 

 grains ; this is true also of that formed on the slopes of Mauna Loa from the summit dis- 

 charges ; but the flow of 1840 from Kilauea at a considerable depth, produced an a-a full of 

 large grains of this mineral. The specific gravity of a specimen from Kilauea is 2.47. The 

 formation of this rough and curious product has already been discussed. 



l Les Isles Ponces, 1788, avant propos, p. 8. ~ Scrope on Volcano!, p. 116. 



MKMOIHS BOST. SOC. NAT. HIST. Vol. I. Pt. 3. 117 



