220 bulletin: museum of comparative zoology. 



fractured specimens reveals the interesting fact that the growth of these plates 

 was by secretion of new layers of bony tissue on both sides of the old, above 

 and below, thus proving that the entire plate was covered by the integument. 

 In no other way can we explain the presence of successive tuberculated layers 

 underneath the external one, nor the regularly laminated structure of the 

 plates as seen in cnjss-seetion. A number of undoubted Arthrodire plates from 

 the Kinderhook near Burlington, Iowa, were collected many years ago by 

 Messrs. Giles, Wachsmuth, and St. John, and are now preserved in the Cam- 

 bridge Museum of Comparative Zoology. 



As if in mimicry of the tuberculated covering of Arthrodires, rugose dermal 

 plates were developed by their Elasmobranch contemporaries in the Kinder- 

 hook, probably through concrescence and fusion of shagreen granules ; and their 

 resemblance to the former is often so close that an examination of the micro- 

 scopical structure is necessary to distinguish them. An example of such a plate, 

 with symmetrical outlines and simulating the tuberculation of Arthrodires, is 

 shown in Plate 5, Fig. 50. The more common form of dermal tubercles. 



Fig. 16. 



Dermal Plate of an undetermined Elasmobranch, in lateral and superior aspects, 

 X f • Kinderhook limestone, Burlington, Iowa. 



however, is acutely or obtusely conical, as exemplified by Petrodus or by the 

 spiniform bodies shown in Plate 5, Figs. 56 and 57. Occasionally bodies are 

 found having the form of elongated eminences, either symmetrical like that 

 shown in Text-figure 16, or abruptly truncated on one side, as if they had been 

 disposed in pairs, and recalling the dermal head plates of Myriacanthus and 

 other Chimaeroids. Many of these tuberculated plates may be referred witli 

 considerable confidence to Chimaeroids, notwithstanding the fact that they are 

 unaccompanied by dental plates. It is a remarkable circumstance that Chi- 

 maeroid jaws, which occur in great profusion in the Middle and Upper De- 

 vonian, are wholly unknown in rocks of Carboniferous age, and Dipnoans are 

 conspicuously absent in the lower members of the same series. An explanation 

 of their sudden disappearance at the close of the Devonian is possibly to be 

 found in the change that took place from shallow to deep water conditions 

 with the resultant migration of littoral forms. 



In Text-figure 17 is shown of twice the natural size a peculiar fossil from one 

 of the " fish-beds " near Burlington, Iowa, stratigraphically near the dividing 

 line between Upper Devonian and typical Kinderhook. It is one of a score or 

 more precisely similar bodies which were collected by St. John, Wachsmuth, 



