84 The Scottish Naturalist. 



Hebrides or anywhere else. Without venturing to express an 

 opinion upon the general question I feel convinced that the great 

 eruptions of lava that formed most of the Inner Hebrides were 

 accompanied by the ejectment of the scoriae, &c, on a very great 

 scale at some, if not all the points I have indicated. 



When we bear in mind that in Skye, Rum, Ardnamurchan and 

 Mull, we find masses of acid lavas whose indifferent fluidity con- 

 fines them to the neighbourhood of the point of outflow, we have 

 an indication that these are the positions of the fissures through 

 which the lavas were ejected, and as deep-seated basic rocks are 

 found in contact with deep-seated acid rocks at the same places, 

 we have further valuable support to the hypothesis. 



Then, in Mull for example, we meet with phenomena that are 

 at least more easily accounted for on the supposition that craters 

 had existed than on any other. The great mountain of Ben More, 

 as well as some of the neighbouring hills, seems to be almost 

 entirely composed of volcanic cinders intersected by numerous 

 sheets and dykes of basalt, while tuffs and scoriaceous rocks are 

 found in patches all over the island ; and near Dunessan is a con- 

 solidated stream of volcanic mud containing numerous fragments 

 of chalk and chalk-flints. The mud seems to be composed of fine 

 volcanic dust, while the chalk and flints were probably blown out 

 of a neighbouring crater ; the chalk is much indurated and closely 

 resembles bits of limestone hardened by heat, such as I have seen 

 among the ejecta of more recent volcanoes. 



Of course, as far as Eigg is concerned, this is merely a digres- 

 sion ; as no one disputes either the nature or origin of the basalts 

 and dolerites of which it is so largely built up. 



A famous section in the basalts at Ardtun in Mull shows that 

 very considerable intervals occurred during the series of eruptions, 

 for here we find quantities of plant-remains embedded in the mud 

 of a lake bottom between sheets of basalt, which indicates a period 

 of rest between the lava-flows of sufficient length to allow denuding 

 agencies to form vegetable soil on which trees of considerable size 

 flourished ; and this must have been no inconsiderable time, as 

 the striae engraved on these rocks during the glacial period have 

 not yet been erased. As the leaf-beds of Mull are repeated 

 several times one above the other with sheets of lava between, it 

 is evident that these periods of rest were frequent as well as pro- 

 longed. Another valuable feature of these leaf-beds is that they 

 supply us with a geological date for the age of these basaltic 



