DEVONIAN FISHES OF IOWA 183 



number of constant characteristics of all the elements thus far 

 brought to light must be admitted to be the true arrangement; 

 that is, the parts must have interacted according to this particu- 

 lar fashion in order to have produced the observed effects, for 

 had they operated after a different fashion they would have 

 produced different effects. 



To anyone who has had the handling of the actual specimens 

 and has experimented with them in the manner indicated, es- 

 cape from the foregoing conclusion seems impossible. The evi- 

 dence for it is not, however, accepted by Dean,* who believes 

 that the normal position of palatal dental plates is indicated 

 by two specimens lying side by side in the slab containing the sin- 

 gle tolerably complete individual described by him in his memoir 

 of 1901. It is assumed by Dean that the juxtaposition of these 

 two palatal plates is natural, not fortuitous owing to post- 

 mortem displacements, as is the case with the dissociated half of 

 the same pavement. The reconstruction proposed by Dean rests 

 entirely upon this unproved assumption, and his arguments in 

 support of it lead from theoretical premises to theoretical conclu- 

 sions. Some of the latter involve if not incredible, at least start- 

 lingly curious features, such as the assignment to Mylostomids 

 of rotary and other complicated jaw movements, the like of 

 which exists nowhere among chordates ; and partly as a corollary 

 to this inference it is suggested that the jaws themselves are not 

 homologous with those of ordinary fishes. The inherent im- 

 probability of the conclusions depending upon Dean's recon- 

 struction is sufficient reason for distrusting the validity of the 

 premises upon which it is based.t A really decisive test of its 

 efficiency may be applied, however, the same in fact -as has al- 

 ready been applied to the restoration proposed by Eastman in 

 1906. This consists simply in fitting the oral surfaces of upper 

 and lower dentition together according to what is assumed to 

 have been their normal position, and observing the mutual cor- 

 respondence of marginal contours and grinding areas, and inter- 

 play of all the parts that are shown by markings to have closed 



* Science, 1907, 26, pp. 46-50. Ibid., 1908, 27, p. 203. 



t"A false conception, when the consequences from it are followed further and 

 further, will sooner or later lead to absurdities and palpable contradictions." — 

 Quoted by Fritz Miiller, in "Facts for Darwin," chap. ii. 



