Correspondence — Prof. E. D. Cope. 573 



action between naturalists in questions of classification and nomen- 

 clatui-e. Scientific truth is involved in the former, and convenience 

 in the latter. 



In the first place I shall be sustained in differing from Mr. Lydek- 

 ker vp-henever he fails to comply with that bulwark of the language 

 of science, the law of priority. Except in the matter of the termina- 

 tion of the famil}'^ name in idee, and its derivation from some genus 

 embraced in it, any name proposed first and accompanied by a defi- 

 nition, must be used in preference to any later name. I take it that 

 this rule applies to all scientific terms whatsoever, which belong to 

 any object, or definite idea abstracted from objects, provided the 

 name be not false in its significance. Hence it will not be proper 

 to yield to mere preferences, as for instance that of Mr. Lydekker, 

 who would rather name orders and suborders from some genus 

 which they contain, when other names have already been proposed 

 in accordance with the rules. 



In the application of this rule to the use of names for divisions of 

 higher rank than families, room for the exercise of some discretion 

 may be found. If a name be applied by its author to a group, there 

 are two ways of learning what the proposer of the name had in view 

 or the idea he intends to express ; or, in other words, what he would 

 include within its limits. One indication is to be found in his diag- 

 nosis ; the other in the contents of described objects which he era- 

 braces in it. I claim, and I find that custom sanctions the claim, 

 that the name should be retained for the division thus indicated, and 

 for nothing else ; and if both diagnosis and content do not represent 

 any natural or tenable division, that the name should be dropped. 

 In insisting on the applicability of diagnosis or content to something 

 real, as the ground of the acceptance of a proposed group and of its 

 name, I of course do not expect completeness in either of the con- 

 ditions. In fact, these characteristics are never to be expected in the 

 early stages of science. But incompleteness is not falsity. An 

 example of a false division with a false diagnosis is to be found in 

 the supposed order of Pachydermata, the name of which is positively 

 rejected, and is not applied to one of those fragments of it which 

 are natural divisions. Another false group is the Theriodonta, which 

 was applied to some South African reptiles of Permian age. The 

 definition, based on the dentition, does not define an order, and is 

 common to more than one family, and the contents of the division 

 agree with the definition. As types with grinding teeth clearly 

 belong to the same order and suborder, there seems to be no place 

 for the name. 



To apply these propositions to the cases found in my papers and 

 criticized by Mr. Lydekker. While the Condylarthra do agree 

 generally with the Hyracoidea, the differences are so important, 

 especiall}' in the form of the ungual phalanges, that I cannot refer 

 them to the same suborder at least. Nor am I justified in applying 

 the term Hyracoidea to the two divisions combined, as they certainly 

 must be, for the group Hyracoidea has already a meaning and cur- 

 rency which it is not wise to attempt to disturb. Should I do so. 



