1226 Note on some Fossil Fishes recently discovered, ete. [December, 
\ 
long, including Pteraspis, Cyathaspis, Scaphaspis and Palzaspis 
is marked off by a remarkable peculiarity of structure from 
every known vertebrate animal. So conspicuous is this charac- 
ter that by its aid we are able to distinguish the smallest frag- 
ment of their bony shields from all other bone whatever. Nota 
trace can be seen in them of the peculiar structure which to the 
anatomist characterizes true bony tissue. The bone cells (la- 
cunz) and the tubules (canaliculi) connecting them are totally 
absent. Nor can the Haversian canals, which are usually present 
in bone, be discovered. So important is this fact that the unwil- 
lingness of some naturalists to admit them among vertebrates 
was natural and defensible. Nor was this unwillingness over- 
come until the discovery of a specimen showing a few scales 
attached to the matrix behind the shield afforded indisputable 
proof of their ichthyic affinity. 
The profound significance of this fact has not yet, it appears to 
me, been fully appreciated by systematists, nor has the fact itself 
received the recognition to which it appears to be entitled. Prob- 
ably the obscurity in which these organisms have been hidden, 
their imperfection and their local and individual scarcity have 
contributed to bring about this result, Yet it is hard to avoid 
the conviction that in this immense anatomical difference there 
may be concealed at least one stone of the bridge that shall some 
day span the gulf now separating the vertebrates and inverte- 
brates, and that in this approximation toa simpler structure we 
may see a link between the highly complex tissue of true bone 
and that simpler and more homogeneous material of which the 
hard parts of animals lower in the scale are composed. It - 
be but one link; others would still be wanting. The relative 
positions of the hæmal and neural elements would present the 
same difficulty, and the riddle propounded by these will perhaps 
be never read, as the soft parts are not likely to be preserved. 
In thus carrying down the remains of fish almost to the re 
of the Upper Silurian rocks, it becomes evident that we em 
seek in some part of our Cambro-silurian series to find yet Ca 
forms. Itis not likely that these are the first that existed. ot 
beds must be searched. Pander’s “ Conodonts” may ye iar 
their place among vertebrates, and fossils older than i - 
come to light before we can say we have found “the mi 
link,” if such ever existed—the Ichthyic Archetype. 
