586 Mn. M. r, woodward on [May 5, 



the base of this structure, the metacone and subsequently the^ 

 hypocone being similarly derived from a backward extension of 

 the base of the primitive dentinal germ. This primitive dentinal 

 germ has, I believe, primarily a somewhat conical form in all cases, 

 and one of the cusps of the adult tooth appears to be the direct 

 continuation of this primitive cone, the remaining cusps being 

 outgrowths usually from its base. It is not customary to find a 

 blunt expanded table-like dentinal germ from which the cusps 

 arise as secondary outgrowths — a condition which, it appears to 

 me, must be necessarily assumed to support Osborn's view that 

 the protocone is primary but retarded aud the paracone its lateral 

 derivative accelerated. 



If it be the case that the paracone in the majority of Mammalia 

 is the direct continuation of tlie primitive dentinal germ, and 

 therefore of the single cone of the protodont mammalian ancestor, 

 then we have the apparent anomaly of this primary cone giving 

 rise, in the majority of forms, to the so-called paracone, i. e. the 

 antero-external cone, while in a few it persists as the so-called 

 protocone (antero-internal cone), a condition which suggests that 

 the usually accepted identification of the cones of the iqyper molars 

 is not in all cases the correct one. 



It may be possible that in the above too much stress is laid on 

 the ontogeny of the molar cusps ; but, on the other hand, do we 

 know suHicient of tlie phylogeny, as deduced from palseontological 

 evidence, to prove that the primitive cone has in all cases been 

 correctly identified in the upper molars ? For though we have, 

 thanks to the researches of Owen (17), Osborn (16), and Marsh (11), 

 knowledge of a great number of Mesozoic mammals, yet the molar 

 teeth found are nearly all lower ones, and but few upper molars (save 

 multituberculate ones) are known until we reach Tertiary times ', 

 when the teeth have assumed forms whose cusps can be more 

 easily homologized with those of living mammals than with the 

 cusp or cusps of the Iloptilian tooth or with that of the ancestral 

 mammal. So that with regard to the evolution of the upper 

 .molars we are almost completely in the dark, for we know of no 

 Triassic or Jurassic protodont upper molars, but three maxillae 

 (I believe) containing tricouodont teeth, and but a few which, 

 according to Osborn, contain trituberculate teeth. 



I have tried to ascertain the exact number of wpjiec jaws of 

 Jurassic mammals possessing tritubercular molars or teeth approxi- 

 mating to that type, bub have been unable to disperse the mystery 

 which seems to envelop them. In England we certainly possess 

 one specimen, which was described by Owen (17) as Peralestes 

 lonc/irostris, and is preserved in the British Museum ; with this 

 Owen associated a lower jaw which is now separated by Lydekker 

 (10) from this form and assigned to Amhlotherium mustelula. Oweii 

 also described fo\ir upper jaws,which he referred to Stylodonpusillus ; 



' Seyeral isolated upper molars are known from the Upper Cretaceous rocks 

 ofN. America; some of these are said to be trituberculate (Osborn, " Mammals 

 of (the Upper Cretaceous Beds," Hull. Amer. Mus. Nat. Hist. 1893, p. 311), 

 notably l)idelpkops, but this, though triangular possesses at least 6 cusps. 



