488 UUM. CiEOLOGT. 



the soil is deep and covered with a thick coating of 

 moory plants on the interior declivities of the hills. 

 Its connexion with the augit rock, is however sufficient 

 to prove that, like this, it reposes immediately on the 

 sandstone; were any proof necessary of the generally 

 overlying position of this rock. 



The connexion between the augit rock and the syenite 

 is of an unexpected nature, and it was not till after a 

 careful examination that I was satisfied respecting it. 

 The subsequent examination of analogous transitions in 

 Mull and in Arran, have entirely removed all doubts 

 on this subject, and confirmed the gradation subsisting 

 between the ordinary traps and the syenitic rocks. It 

 was before remarked that the augit rock assumed a 

 stratified form on the shore of Harris, and was disposed 

 in thin beds alternating with a rock resembling basalt, 

 but of which augit formed a constituent part. Suddenly, 

 and without well seeing how this stratified rock dis- 

 appears, it is found replaced by great masses of syenite, 

 having at this particular place no tendency, or one 

 scarcely visible, to that bedded disposition for which 

 it is so remarkable in the cliffs near Bridianoch. By 

 carefully and repeatedly following the stratified rock, it 

 is at length discovered that there is a regular transition 

 from the one to the other ; the augit rock at first be- 

 coming finer grained, and acquiring a greater proportion 

 of felspar, which gradually increasing, ultimately produces 

 the syenite. Occasionally, the transition is a little more 

 rapid, and although no decided line will be perceived 

 on breaking the rocks at the place of transition, the 

 different modes in which the several sorts weather, point 

 out the place where the change is effected. In no 

 case however is there any discontinuity or interruption 

 in the masses, which, in their external forms, pass into 

 each other by the most insensible gradation. This change 

 in the forms is even more remarkable than that in the com- 



