THE TRAP ISLANDS. 6l 



Two other kinds of error will be found to interfere with 

 the correct determination of the directions of the strata. 

 The first of these arises from an undulation of the line of 

 bearing, which although sufficiently constant where a 

 large space is assumed, with certain decided exceptions 

 that will here appear, is found to vacillate within certain 

 limits ; rarely amounting to a point of the compass, and 

 never exceeding two. The second cause of error, is an 

 uncertainty arising from the unequal variations of the 

 magnetic needle. With a sufficient degree of labour 

 and time it might have been possible to determine that 

 variation at every point ; as well as the actual deviations 

 of the bearings and dips, all of which have been noted in 

 a general manner. But such observations, could they 

 even be perfected, are of no value unless they conduce to 

 some general object. That for which they have, both here 

 and elsewhere, been brought into notice, will be as effec- 

 tually accomplished by taking them within the limits of 

 error now acknowledged, as if every dip and bearing had 

 been determined to a degree. In laying down the bear- 

 ings, the variation of the needle has every where been 

 assumed at 23 west ; although this, as I formerly remarked, 

 is by no means a rate universally applicable. 



In the tract which includes the northern division of the 

 Trap isles, there are examples both of the primary and se- 

 condary strata ; the former being limited to gneiss and 

 primary sandstone, using these terms in a comprehensive 

 sense, the latter including a limestone, here distinguished 

 by the name of gryphite limestone, and a set of strata 

 which may, without much risk of error, be classed with 

 the lias and those arenaceous rocks that immediately fol- 

 low it. 



It has been fully shown that both the gneiss and the red 

 sandstone of Sky are prolonged in a north-easterly direc- 

 tion to the mainland, with which they are therefore con- 

 nected on this line of their bearing. They are moreover 

 connected with the same strata by a system of alternation ; 



