pale sandstone to which the hills of Hoy and
the rocks of Duncansbay Head and of Tarbet Ness belong. Further, in the
second place, while the upper or Holoptychian formation is found
_directly_ overlying that of the Coccosteus in only one
locality,--Moray,--we find it directly overlying that of the Cephalaspis
in _two_ widely separated localities;--in the vast band of Old Red which
runs diagonally across the island from sea to sea, parallel to the
Grampian chain, and in the immensely developed Red Sandstones of England
and Wales. And it is of course more probable that the two corroborative
instances should represent the natural succession of the formations, and
the single instance the accidental gap in the scale consequent on the
missing formation, than that, _vice versa_, the solitary instance should
represent the natural succession, while the two mutually corroborative
ones should represent, in localities widely apart, the mere accident of
the gap. But, in the third place, I attach more weight to a conclusion
founded on the positive character of the groups of organic remains by
which the three great formations of the Old Red System are
characterized, than I do to either of these considerations. The
organisms of the Cephalaspian deposits differ _generically_, and in
their whole aspect, from both those of the Coccostean and Holoptychian
formations; whereas the organisms of the Coccostean formations, while
they resemble generically and in the group those of the Holoptychian
one, mainly differ from them _specifically_. The extreme _generic_
difference in the one case argues evidently a great difference in
_condition_,--the lesser specific difference in the other, a great
difference in point _of time_. The Cephalaspian formation might, as a
fresh water formation, be nearly contemporary with either of the other
two, or, as seems more probable, interposed between them; while they
themselves, on the other hand, generically similar and decidedly marine
in their character, must have been so widely separated in time, that all
the species of the lower group became extinct ere those of the upper one
had been ushered into being. And such are some of the considerations
that still lead me, notwithstanding the failure of previous evidence, to
hold, at least, provisionally, that our Scottish flagstones to the north
of the Grampians occupy a lower horizon than our Scottish tilestones to
the south. It must, however, be stated, on the other