Plant Invasion on Lava Flows. 17 



pahoehoe. The aa assumes the position of a winding river through 

 the pahoehoe. In places the pahoehoe has flowed around portions 

 of the aa in an irregular manner, leaving sunken islands of aa 

 varying from a few feet to half an acre in area. From the close 

 interrelations between the two there is no doubt that both belong 

 to the same flow. The pahoehoe is jet black and new looking, 

 while the aa has a chocolate tinge, it also appearing very new as 

 compared to adjacent flows. 



After several hours spent on the flow it became apparent that, 

 with the exception of the lower cryptogams, most of the vegetation 

 was supported on the pahoehoe and not on the aa. As this was 

 contrary to what I had expected and had heard generally ex- 

 pressed on these islands, I gave particular attention to this point, 

 and also directed my guide to call my attention to any plants he 

 might see on the aa. A portion of the aa stream was followed for 

 a considerable distance down the centre, but nothing was obtained 

 save a liberal supply of cuts and bruises. This scarcity of plants was 

 even true for the small sunken areas of aa, which of all places one 

 would think would be admirable traps for catching seeds and 

 spores. The aa portions of the flow are often white with a certain 

 lichen, and a closer search reveals an occasional moss, but there 

 are no ferns or phanerogams, except on the contact line with the 

 pahoehoe. 



The plants which occur on the pahoehoe are to be found grow- 

 ing in the numerous small cracks which cross the flow in various 

 directions, especially where the flow is at all billowy, and between 

 the folds of the ropy lava. The smoother portions are entirely 

 bare of any vegetation. Such cracks act as riffles to catch either 

 disintegrated particles of the flow itself, or dust and other refuse 

 which may be blown over the smooth portions from the outside. 

 At one place I noticed several hundred dead shells of Eulota 

 ntnilaris being blown across a portion of the flow, many of them 

 being caught in the cracks. Such soil is stopped on the edge of 

 an aa flow, while apparently not enough accumulates by the 

 weathering of the flow itself in this period of time to fill its 

 smallest spaces. 



The following plants were observed scattered here and there 

 in the < ra< k^ 011 the pahoehoe, but they were in no case in suffi 



l-wii'S It P H. M. Vol.. V No. I.— a. 



