212 bulletin: museum of comparative zoology. 



the size of the upper dental plates which must have accompanied ik 

 during life can be predicted with entire accuracy. 



Experience having shown that all of the dissociated dental elements 

 now known, upper and lower, exhibit among themselves practically 

 uniform dimensions and uniform conformation, one is entitled to con- 

 clude therefrom that they represent average-sized individuals, and that 

 the elements were arranged after an invariable pattern. For, supposing 

 their disposition to have been inconstant, we should be at a loss to 

 account for their raai'ked regularity of form and proportion, and similar 

 indications of wear. Hence any theoretical reconstruction of the den- 

 tition, whether based upon detached specimens or upon the evidence of 

 naturally associated parts, must satisfy the test of totality. It must 

 apply universally, not only to such plates as are known to have 

 belonged to a single individual, but to all those that have been found in 

 the detached condition as well ; it must be compatible with all their 

 essential features, and be negatived by none of them. 



It may be that only one, or more than one theoretical reconstruction 

 of the dentition is competent to explain all the observed facts. As 

 between two rival hypotheses, that one may be regarded as tlie more 

 plausible which is mechanically simple, free from anomalous supposi- 

 tions, and in harmony with analogy. An hypothesis which is mechan- 

 ically complicated, presupposes anomalous conditions, and violates 

 analogy, is less worthy of credence. For in so far as it depends upon 

 the assumption of the unique, of something for which nature affords no 

 parallel, it becomes improbable ; and the improbable is always to be 

 distrusted. Speaking broadly, any hypothesis whatsoever has the ele- 

 ments of trustworthiness, provided it can be shown to agree with a 

 number of divei'se facts. The greater number of diverse facts with 

 which it agrees, the more completely can it be verified. When many 

 circumstances point toward a single conclusion, the chances of that 

 conclusion being coirect are enormously increased with each additional 

 fiivoring circumstance. They might even be supposed to increase in 

 geometrical rather than in arithmetical ratio. Finally, an hypothesis 

 that is found to agree entirely with observed facts cannot but be believed 

 to be true. It will be instructive to inquire how far either of the two 

 extant interpi-etations of Mylostomid dentition are in accordance with 

 observed facts. 



Newberry's Views. — Our earliest information regarding the Mylos- 

 tomid type of dentition is due to the zeal and acumen of Professor J. S. 

 Newberry, who described the constituent elements of the type species, 



