THE SCHISTOSE ISLES. 287 



mainland also, in the position of the strata at Degnish 

 point ; but the occurrence of the trap at this place puts 

 a stop to all further investigation in this quarter. 



Throughout the whole extent of these varying di- 

 rections, it is obvious that the changes correspond with 

 those of the Quartz isles and of the Slate isles; and 

 that a general parallelism is thus maintained under all 

 the irregularities, by every member of the group of the 

 Schistose isles at large, and by the strata of whatever 

 nature that form the adjoining continent. 



The variations which take place in the dip of the 

 chlorite strata throughout their extent, are considerable, 

 and merit a somewhat detailed consideration. Com- 

 mencing at West Tarbet Loch, which, in a general 

 sense, may be considered as the south-eastern boundary 

 of the series, the dip is towards the west; in which 

 respect it corresponds with that of Gigha already de- 

 scribed. The same regularity is not however maintained 

 on the prolongations of the strata, as may be discovered 

 by examining the western shore of south Loch Fine. 

 Here, for a certain space from Barmore northwards, 

 the dip is frequently reversed; the strata thus often 

 occupying the vertical line, and considerable irregularity 

 being the result. But from Strandore, on this section, 

 where the strata are vertical, to Loch Gilp, the north- 

 western dip is regular ; or, if there are any irregularities, 

 they are so limited as not to be easily discovered. On 

 the western section, on the contrary, both in South 

 and North Knapdale, irregularities similar to those near 

 Barmore, are found ; although by far the prevailing 

 dip of the whole series is still! to the westward. These 

 irregularities, like the former, consist in occasional vacil- 

 lations from the eastern to the western dip, accompanied 

 by vertical intermediate portions. Notwithstanding these 

 vacillations however, the regularity of the strata and 

 the parallelism and evenness of their alternations are 

 never changed ; nor are the reversals ever attended 



3 



