its longer
diameter. I have ascertained further, that this longer diameter was
equal to the shorter diameter of the creature's frontal buckler,
measured across about two thirds of its entire length from the nape; and
that a transverse diameter of ten inches at this point was associated in
the buckler with a longitudinal diameter of fourteen inches from the
nape to the snout. Thus five inches along the nail represent fourteen
inches along the occipital shield. The proportion, however, which the
latter bore to the entire body in this genus has still to be determined.
The corresponding frontal shield in the Coccosteus was equal to about
one-fifth the creature's entire length, and in the Osteolepis and
Diplopterus, to nearly one-seventh its length; while the length of the
_Glyptolepis leptopterus_, a fish of the same family as the Asterolepis,
was about five and a half times that of its occipital shield. If the
Asterolepis was formed in the proportions of the Diplopterus, the
ancient individual to which this nail-like bone belonged must have been
about eight feet two inches in length; but if moulded, as it more
probably was, in the proportions of the Glyptolepis, only six feet five
inches. All the Coelacanths, however, were exceedingly massive in
proportion to their length; they were fish built in the square,
muscular, thick-set, Dirk-Hatterick and Balfour-of-Burley style; and of
the Russian specimens, some of the larger bones must have belonged to
individuals of from twice to thrice the length of the Stromness one.

Passing upwards along the strata, step by step, as along a fallen stair,
each stratum presenting a nearly perpendicular front, but losing, in the
downward slant of the _tread_, as a carpenter would say, the height
attained in the _rise_, I came, about a quarter of a mile farther to the
west, and several hundred feet higher in the formation, upon a fissile
dark-colored bed, largely charged with ichthyolites. The fish I found
ranged in three layers,--the lower layer consisting almost exclusively
of Dipterians, chiefly Osteolepides; the middle layer, of Acanthodians,
of the genera Cheiracanthus and Diplacanthus; and the upper layer, of
Cephalaspides, mostly of one species, the _Coccosteus decipiens_. I
found exactly the same arrangement in a bed considerably higher in the
system, which occurs a full mile farther on,--the Dipterians at the
bottom, the Acanthodians in the middle, and the Cephalaspides atop; and
was