The species extreme in character are not the oldest, or the most recent; nor are those which are intermediate in character intermediate in age.

But supposing for an instant, in this and other such cases, that the record of the first appearance and disappearance of the species was perfect, we have no reason to believe that forms successively produced necessarily endure for corresponding lengths of time.

A very ancient form might occasionally last much longer than a form elsewhere subsequently produced, especially in the case of terrestrial productions inhabiting separated districts" (pages 334-5).

The same words occur in the later edition of the "Origin" (Edition VI., page 306.) All these new discoveries show how imperfect the discovered series is, which Falconer thought years ago was nearly perfect.
I will send to-day or to-morrow two articles by Asa Gray.

The longer one (now not finally corrected) will come out in the October "Atlantic Monthly," and they can be got at Trubner's.