Note on a Basalt Tree Cast. 143 



for the lava to the mould we have two alternatives ; either the 

 first flow could not have completely covered the tree leaving some 

 portion projecting above the surface as exemplified at Kilauea, or 

 else the tree was completely covered by perhaps many feet of 

 basalt, which has in the course of time through denudation been 

 gradually removed until the mould was reached and rendered 

 accessible to the succeeding flow. The latter view does not seem 

 probable, because even provided that the charred remains of the 

 tree were removed from the cavity the products of deconi position 

 and detritus from the surface would have to a large extent, if 

 not completely, tilled it and prevented anything like a perfect 

 cast being formed. Besides this, by the time the mould was 

 reached decomposition of the basalt forming it would in all 

 probability have advanced so far as to destroy the clearness of 

 the mould. The tirst alteinative seems most satisfactory and 

 presuming that the tree occupied a horizontal position it would 

 mean that the lava sheet could only have attained a thickness of 

 about 2 feet at the point where the tree was buried, and even if 

 an upright position had been maintained its thickness could not 

 have been more than 10 feet. INIolten lava streams obey the 

 laws of all liquid bodies and, therefore, in their movements 

 resemble water-courses, following valleys and forming streams of 

 sometimes great thickness, where the country is hilly or 

 mountainous, and flooding it when it assumes a slightly 

 undulating or level character. The tree may then have been 

 growing in comparatively flat country, or on an eminence when 

 it was borne down upon by a flow of molten lava, and unable to 

 withstand the weight of the advancing mass was forced down 

 before it breaking its roots off short by the stem and then rapidly 

 being almost covered by the flow. The length of time which 

 then elapsed, during which the mould became freed of the 

 carbonized remains of the tree, and the advent of the succeeding 

 flow in the absence of authentic data are problems difl&cult to 

 explain. It may have happened that the tree remains slowly 

 smouldered away, a supposition entailing a coolness of the basalt 

 in the immediate vicinity of the mould, which otherwise would 

 have rapidly closed. We have seen that lava in a highly molten 

 state thrown on to tree branches has done no more than scorch 

 the bai'k, and also that tops of trees which have been enveloped 



