ing down of the very thin ones of _horn_
which cover fish such as the carp or salmon, was, that in the massier
_slates_, the sides, or _cover_,--nicely bevelled, in order to preserve
an equability of thickness throughout,--were so adjusted, that two
scales at their edges, where they lay the one over the other, were not
thicker than one scale at its centre. Even in the other ganoids, their
contemporaries, such as the Osteolepis and Diplopterus, where the
scales were ranged more in the tile fashion, side by side, there was,
with much ingenious carpentry in the fitting, a general simplicity of
form. It would almost appear, however, that ere the ganoid order reached
the times of the Weald, the simple forms had been exhausted, and that
nature, abhorring repetition, and ever stamping upon the scales some
specific characteristic of the creature that bore them, was obliged to
have recourse to forms of a more complex and involved outline. These
latter-day scales send out nail-like spikes laterally and atop, to lay
hold upon their neighbors, and exhibit in their undersides grooves that
accommodated the nails sent out, in turn, by their neighbors, to lay
hold upon _them_. Their forms, too, are indescribably various and
fantastic. It seems curious enough, that immediately after this
extremely _artificial_ state of things, if I may so speak, the two
prevailing orders of the fish of the present day, the Cycloids and
Ctenoids, should have been ushered upon the scene, and more than the
original simplicity of scale restored. There took place a sudden
reaction, from the fantastic and the complex to the simple and the
plain.

It is further worthy of notice, that though many of the ganoid scales of
the Secondary systems, including those of the Wealden, glitter as
brightly in burnished enamel as the more splendent scales of the Old Red
Sandstone and Coal Measures, there is a curious peculiarity exhibited in
the structure of many of the older scales of the highly enamelled class,
which, so far as I have yet seen, does not extend beyond the Palaeozoic
period. The outer layer of the scale, which lies over a middle layer of
a cellular cancellated structure, and corresponds, apparently, with that
scarf-skin which in the human subject overlies the _rete mucosum_, is
thickly set over with microscopic pores, funnel-shaped in the transverse
section, and which, examined by a good glass, in the horizontal one
resemble the puncturings of a sieve. The M