
506 THE. AMERICAN NATURALIST. | (Vor. XXXVI. 
sense visual organs, or, in fact, anything but mere muscular 
impressions ? 
But these are trite criticisms in comparison with what might be 
said respecting Patten’s remarkable declaration that ostracoderms 
possessed many pairs of jointed appendages. To refute this is to 
nullify the author’s sole new claim that these creatures are “ genet- 
ically related to arthropods, or that, through changes in structure and 
function, one type has been derived from the other"; for all other 
of the well-known resemblances may be explained as due to mimicry, 
or to incidental parallelism dependent on environment. 
We have already characterized this assertion of Patten's as a sur- 
mise; we will go further, and say that it is not only an unproved 
hypothesis, but one that is absolutely unsupported by any specimens 
that have yet been described. If examples of Cephalaspis have 
been found which display “a fringe of 25 to 3o pairs of jointed and 
movable appendages along the ventral margin of the trunk," the 
present reviewer for one will confess that he has never seen nor 
heard of them. As for * Lindstróm's important discovery of an 
appendage in Cyathaspis," this detached and unjointed fragment is 
probably to be interpreted as a spine or cornu. But Professor 
Patten does not stop here: he points to a series of marginal openings 
in the shield of Pteraspis, Cyathaspis, and Tremataspis, commonly 
regarded as branchial, and declares that they “must” have served for 
the attachment of appendages. He surmises that in Tremataspis 
the latter ** decreased in size from before backwards, and were possibly 
too delicate to be well preserved in a fossil condition." Fearful lest 
our rude touch should annihilate these frail conjectural organs, we 
will simply refer those interested to Dr. Traquair's comparison of 
Tremataspis with Birkenia, in which the branchial openings (com- 
monly so-called) are disposed relatively as in the skate and shark.’ 
A comparison, also, of the anterior ventral plates of Tremataspis 
with the very similar ones of Drepanaspis, in our opinion would 
have been far more instructive than our esteemed friend's attempt 
to homologize certain of them with the jaws of an arthropod. 
"The position of the mouth in Tremataspis we shall believe, until 
the contrary is proved, to have been correctly determined by Rohon, 
who places it immediately behind the rim of the head-shield, as in 
other ostracoderms where the oral plates are satisfactorily known. 
Patten, however, basing his conclusions on the solitary example 
: studied by Rohon, challenges the latter's interpretation, and would 
1 Trans. Roy. Soc. Edinb., vol. xxxix (1899), pt. iii, p. 859. 


