THE AMERICAN NATURALIST. [Vol. XXXVI I. 



There can be no doubt that the fringing processes projected 

 freely from the ventral margin, and that they were freely mov- 

 able forward and backward, and to and from the median line. 



The conditions we have described are not less extraordinary 

 than the fact that such conspicuous structures should have 

 remained so long practically unnoticed. Lankester does not 

 discuss their possible signification in his text, but merely intro- 

 duces them into a very diagrammatic cross section as plates pro- 

 truding like bilge keels away from the body. Many of the 

 newer specimens of Ccphalaspis miur/iisoni in the British 

 Museum, as well as some of the older type specimens, show in 

 the clearest manner that the fringing processes are articulated 

 to the ventral margins of the body and are not artificial folds 

 made by pressing together the margins of the dorsal and ventral 

 walls (Fig. 7). 



That the whole group of ostracoderms was provided with a 

 series of fringing plates similar to those of Cephalaspis is very 

 probable, for a series of fringing plates are known to exist in 

 Tremataspis, and I have found indications of fringing processes 

 in the trunk of a fine specimen of Pterichthys preserved in the 

 geological collection of McGill University. 



Morphology of Vertebrate Appendages. — Organs so widely 

 distributed in a primitive group of animals as the fringing proc- 

 esses are, must have great morphological significance. But 

 while there is little doubt that they are the antecedents of the 

 lateral fold of vertebrates, for no other structures so clearly 

 reproduce in size, position and function the hypothetical folds 

 from which the paired fins are supposed to take their origin, 

 that fact does not help us to determine the morphological 

 significance of the fringing plates themselves. 



The fact that the fringing plates are marked with the same 



comparable with internal fin rays. On the contrary, their orna- 

 mentation, shape, and mode of articulation indicate that they 

 are independent, segmental structures. It is difficult to inter- 

 pret such structures as anything else than appendages, having 

 the same significance as the rudimentary abdominal appendages 

 of arthropods. 



