
ON THE STRUCTURE AND CLASSIFICATION OF 
THE TREMATASPID.;E.! 
WILLIAM PATTEN. 
OF all the families usually included in the problematical and 
heterogeneous group of animals called the “ ostracoderms,”’ none 
is more interesting to the morphologist than the Tremataspidz, 
as the little known about them shows they possessed a most 
extraordinary structure, unlike in many respects that of any other 
group of animals. While the character of the trunk scales, of 
the orbits and other sensory openings, the minute structure of 
the shell, and the presence of the newly discovered lateral-line 
system clearly indicate the affinity of Tremataspis with Pteras- 
pis, Cephalaspis, and Pterichthys, and through them with the 
true vertebrates, other features, such as the general shape ‘of 
the shield and its more superficial texture, which have long 
been a source of perplexity to the paleontologist, clearly 
point to the affinity of Tremataspis with arthropods like Limu- 
lus, Apus, and the trilobites. 
The importance of the Tremataspide to the morphologist 
also lies in the fact that while the specimens are rare and more 
. or less fragmentary, they are usually well preserved and give 
fair promise that ultimately we shall be able to decipher in 
detail the structure of all their hard parts. This knowledge 
will certainly throw much light on the morphology of the whole 
group of ostracoderms, and may afford decisive evidence of the 
genetic relationship between the vertebrates and invertebrates. 
en, therefore, through the generosity of the administra- 
tion of Dartmouth College, I was granted a half-year's leave 
of absence, I decided to make as thorough an investigation of 
the ostracoderms as my time and means would allow, with the 
special object of determining whether any evidence could be 
1 This paper is an abstract of one about to be published in the Memoirs of the 
Imperial Academy of Sciences of St. Petersburg. 
379 
