Nevertheless this impression is partly false.

Mr.

Croll, in an interesting paper, remarks that we do not err "in forming too great a conception of the length of geological periods," but in estimating them by years.