DEVONIAN FISHES OF IOWA 247 



it was necessary to postulate the existence of their ancestors 

 dnring the Devonian, although their remains had escaped notice 

 both in Europe and North America. Frequently it happens in 

 natural as well as in physical science that, through the exer- 

 cise of trained imagination, predictions can be made with such 

 confidence that our faith in them amounts almost to a certainty, 

 even though there appear to be little chance of ultimate demon- 

 stration of the truth;* but in all such cases it is gratifying when 

 well-founded forecasts happen to become verified by fresh dis- 

 coveries. Smith Woodward's recognition of a Devonian Coel- 

 acanth in the new form brought to light by von Koenen is a 

 good instance in point. Another instance, even more striking 

 than the first, is found in the fortunate discovery by Dr. Stuart 

 Weller of a completely developed Coelacanth, immediately to be 

 described under the name of C. welleri, at the base of the Kinder- 

 hook limestone near Burlington, Iowa. These discoveries com- 

 pel us to project the origin of the family backward in point of 

 time to an earlier period than was at first thought necessary. 



Coelacanthus welleri, Eastman. 

 (Plate III, Fig. 7; Text-fig. 36) 

 1908. Coelacanthus welleri C. R. Eastman, Journ. Geol. 16, p. 358, text-fig. 1. 



Holotype a somewhat imperfect fish, the total length of which 

 to the base of the caudal fin is about 19 cm, or a little more 

 than three times the length of the head with opercular appar- 



*The necessity for postulates of this nature in palaeontology is too obvious, and 

 the practice of making them too common, for illustrations not to suggest them- 

 selves readily to all well informed persons. We may, however, be allowed to cite 

 an excellent recent example of this art of visualizing undiscovered prototypes of 

 fossil forms, which is to be found in Dr. Bashford Dean's recently published 

 "Notes on Acanthodian Sharks" (Amer. Journ. Anat. 1907, 7, p. 220). Having 

 reached the conclusion that Acanthodian sharks "have passed through a stage 

 which is best represented by the Cladoselachian," the author seeks to answer the 

 inquiry why it is that the more specialized group of Acanthodians is known from 

 an earlier horizon than the less specialized? The situation is met by framing the 

 following hypothesis: 



"This is in truth a question which can be answered only by the time-worn 

 appeal to the defectiveness of the palaeontological record, noting especially in this 

 regard that the soft structures of the Cladoselachians would be less apt to be pre- 

 served than the hard structures of the Acanthodians. We may, however, safely 

 predict that from the earliest Acanthodian horizon there will be discovered forms 

 which will represent the ancestors of all the early groups of sharks. And we may 

 predict with the same degree of security that these forms will be found to picture 

 the Cladoselachian in essential characters. For the Acanthodians, as we at pres- 

 ent know them, are obviously too specialized to have represented the ancestors of 

 the line of Cladoselachians." (cf. supra, pp. 61, 99.) 



