May 23, 1913] 



SCIENCE 



791 



found erroneous and infinite recrimination 

 and heartburning wlien the work of earlier au- 

 thors is set aside or ignored. These troubles 

 we shall have with us always; but perhaps 

 their amount might be reduced if an inter- 

 mediate course were adopted. 



The earlier type may be a specimen show- 

 ing unmistakable ordinal, family or generic 

 characters, but not adequate as a specific type. 

 Let it stand so. Do not set it aside as " in- 

 determinate," but specify the extent to which 

 it is determinable. It can remain in the liter- 

 ature and be included if desirable in faunal 

 lists, but additional material should not be re- 

 ferred to it unless the new specimens be topo- 

 types, i. e., from the same locality and the 

 same geological level, so far as these are re- 

 corded or can be safely inferred from the 

 literature, unpublished notes or labels or the 

 appearance of the specimen. If it has valid 

 generic characters a genus founded upon it 

 is valid, and other species may be referred to 

 it; if it has family characters but no distinc- 

 tive generic characters, a family name founded 

 on the genus is valid, but no subsequent gen- 

 era are to be synonymized with it except when 

 species of those genera are known to occur in 

 the locality and geological horizon of the older 

 genotype species. In illustration a few cases 

 may be cited: 



1. Anchippodus riparius Leidy 1868, type a 

 lower molar from the " Miocene " ( ? Oligo- 

 eene) of New Jersey. Type of the family 

 Anchippodontidse Gill 18T2, referred to the 

 order Tillodontia Marsh 1875. Leidy referred 

 to this genus and species in 1873, his Trogosus 

 castoridens 1871 based on a lower jaw from 

 the Middle Eocene (Bridger formation) of 

 Wyoming and to the same genus Marsh's 

 Palwosyops minor 1871, based on a lower 

 molar. Marsh, subsequently obtaining com- 

 plete skulls and skeletons of related animals, 

 accepted Leidy's genus Anchippodus, described 

 a new genus Tilloiherium with three new 

 species, and based upon it the family Til- 

 lotheridse which he made typical of the order 

 Tillodontia. 



No topotypes of Anchippodus riparius are 

 .known. Subsequent authors have either fol- 



lowed Marsh in ignoring Gill's name, while 

 accepting Leidy's identification of Anchip- 

 podus with Trogosus, and considering Til- 

 loiherium as distinct from the latter, or they 

 have used AnchippodontidiB as the family 

 name, while deriving all the characters of the 

 group from Bridger materials. 



The result is that the faunal lists record in 

 the New Jersey " Miocene " along with a 

 known Oligocene mammalian fauna (Oceno- 

 pus, Entelodon, Protapirus) a genus and 

 species of the Middle Eocene fauna, while the 

 western collections make it reasonably certain 

 that in those regions the family and order dis- 

 appeared with the Middle Eocene. Were this 

 conclusion supported by real evidence, it 

 would lead to some interesting corollaries as to 

 migration and survival. In fact it is quite 

 misleading. The type of Anchippodus riparius 

 is inadequate for specific or generic compari- 

 son, and doubtfully adequate for family or 

 ordinal comparison. It is very improbable 

 that it is congeneric with Trogosus, hardly 

 possible that it is co-specific with T. castori- 

 dens, so far as one may judge from the associ- 

 ated fauna in absence of generic or specific 

 characters in the type specimen. Gill's family 

 characters were drawn from Trogosus, and 

 since it is doubtful whether this genus belongs 

 in the same family with Anchippodus, his 

 family should be held as doubtfully synony- 

 mous with Tillotheriida3, both names to be re- 

 tained, but the former as " ? Anchippodontida3 

 Gill, fam. indet." 



2. Eippodon Leidy is the first genus of 

 three-toed horses described from this country. 

 The type is H. speciosus, based upon a lower 

 molar tooth. Leidy subsequently referred to 

 the species upper teeth, etc., which he consid- 

 ered congeneric with the older European 

 genus Hipparion. On the basis of these and 

 other referred specimens the species was held 

 valid and the genus a synonym of Hipparion 

 until Gidley revised the three-toed horses in 

 1907. No topotypes were or are known. Gid- 

 ley set aside both genus and species as inde- 

 terminate. I subsequently identified and lo- 

 cated the type specimen which had been miss- 

 ing, and after making a fairly careful com- 



