THE TRAP ISLANDS. 75 



gneiss ; and in the latter, limestone alone occurs. Should 

 it even be found in a few partial spots it would only 

 confirm the view to be deduced from its equally partial 

 occurrence in Inch Kenneth. That conclusion is, that this 

 apparently important member of the secondary strata, may 

 occasionally be absent even for a considerable space ; and 

 that it is not therefore necessarily the first in order of these 

 rocks. On the contrary, it either occurs in a limited 

 manner, or may be altogether absent. Thus some of the 

 higher members of the secondary strata repose on the 

 primary ; and in this particular case they repose imme- 

 diately on that gneiss which follows the granite. The 

 geological series here, is therefore, granite, gneiss, organic 

 strata with coal ; an arrangement which forms a conspi- 

 cuous exception to the order commonly assumed as preva- 

 lent in nature. It is unnecessary now to recapitulate the 

 circumstances in the structure of the northern division of 

 the Trap islands which confirm this view ; but the reader 

 cannot fail to recollect, that in those also, no secondary 

 red sandstone occurs ; but that the upper members of 

 the secondary strata are found in immediate contact 

 with gneiss and other primary rocks. The extent of this 

 deficiency is therefore considerable ; and if it is only by the 

 absolute magnitude of an exception that its value is to be 

 estimated, geologists will not refuse to consider this one 

 as of an important nature. 



It is difficult to discover any general principle by which 

 to compare the unstratified rocks. The causes of this 

 difficulty must already have appeared in the observations 

 which have accompanied the descriptions of these masses 

 in the several islands where they occur. Although many 

 different substances are found among them, these are all 

 so blended, and united by such imperceptible gradations, 

 that the posteriority or priority of the one or the other 

 cannot be discovered, even if such differences in relative 

 time should actually exist between them. The difficulty, 



