178 Bulletin Wisconsin National History society. [Vol. 9, No. 4. 



uralists that a considerable percentage of cases where several 

 specimens have been used as types, subsequent study has shown 

 these specimens to belong to different species and in some cases to 

 different genera." 



While this fact is in most cases an objection, in palaeontology 

 it is often necessary to describe new species from very fragmen- 

 tary material and to obtain any clear ideas as to just what a per- 

 fect specimen would be, it is always customary to use several or 

 many specimens to get that idea so that the types where they 

 can be obtained must in almost all cases consist of several speci- 

 mens, therefore a series of types ; and while it is doubtful if 

 many species have been defined from a single specimen although 

 that must be the case where the one specimen only is known as 

 it sometimes is, we can readily see that a number of types become 

 a necessity and therefore the type specimens of any natural his- 

 tory object, that is the actual material on which published descrip- 

 tions and figures have been based, are the only ones that should 

 be recognized as the true types and that no other term than that 

 of type should be used at any time. Such objects once lost or 

 destroyed, any replacement is impossible and however imperfect 

 or fragmentary the type may be. or however perfect or better a 

 more recent example of the same creature may be, it cannot serve 

 to replace the functions of the first, as the type specimen is the 

 only one that can be the basis of comparison for all time. 



The term type as employed by the Geological Department of 

 the American Museum, includes not only those specimens actually 

 used bv the author in the original description of a species, but also 

 those specimens which have been used by the same author in the 

 further elucidation of the species in subsequent publications ; the 

 latter therefore become a pari of the type specimens. The original 

 types may or may not have been illustrated in connection with the 



