THE TRAP ISLANDS. 63 



disposed in an irregular cavity resembling that which has 

 been termed a basin ; examples of which are of frequent 

 occurrence among the secondary rocks.* That conclusion 

 may perhaps be also deduced from its absence in Rasay, 

 where the same shale and calcareous sandstone that follow 

 the lias in Sky, appear to repose immediately on the red 

 sandstone. Yet even this appearance admits of another 

 explanation ; since the gryphite limestone may, even in 

 Rasay, lie beneath the uppermost strata, although the 

 immediate contact of these with the red sandstone only, 

 should be visible. This solution depends on the mode of 

 arrangement of the strata in basin-shaped formations ; and 

 as it cannot be rendered easily intelligible in words I have 

 added a section for its illustration :-f the basin in this case 

 will be common both to the gryphite limestone and the 

 strata that follow. The view formerly presented in Scalpa, 

 of its probable division into separate cavities on the irre- 

 gular surface of the sandstone, will tend to illustrate and 

 confirm this explanation. 



The last of the secondary strata, are those which follow 

 the gryphite limestone, which I have referred to the lias 

 series, and to those strata which, in other situations, are 

 known to succeed it. The state of dispersion in which 

 these are generally here found, prevents, in almost every 

 place, any other comparison than that which may be 

 founded on their similarity of composition and on the 

 resemblance of the organic remains they contain. In 

 describing Sky and Rasay, I have been already obliged 

 to draw this comparison as far as these two islands are 

 concerned, and I need not therefore recur to it. In the case 

 of the other islands it was more slightly touched on. The 

 occurrence of the lias among these islands, will probably 

 however be considered an interesting circumstance ; since 



* Although this limestone should be considered a portion of the lias 

 series, it will not affect this reasoning, which would be equally applicable 

 to it as the lowest bed of that deposit. 



t Plate XXIII. fig. 4. 



