Teeth in the Centctidaj. 535 



Since the united para- and mctacone extends to the inner 

 margin of the tooth, and tlie " protocone," from Potamocjah 

 onwards, appears more and more coalesced with the former, it 

 ia probable that the continually diminishing " |)rotoeone" has 

 gradually been absorbed by the para- + mctacone. The state 

 of the case in the molars of Ericulus and Echinops is similar 

 to that which is seen in Centetes ; there is no occasion here 

 to go further into details. — In Orijzoryctrs the protocone is 

 still of relatively considerable size, bicuspid in the case of tho 

 two anterior molars and in P. 1, most distinctly so in the 

 latter; the posterior cusp, wliich appears unusually little 

 developed in M. 1 and M. 2, will have to be interj)reted in the 

 same manner as in Solenodon. — In ^[icrogale Dohsoni also 

 the feebly developed talon is bicuspid. — In Limnogale it is 

 even trieusi)id in the case of perfect teeth. — The maximum of 

 reduction seen in Centetidaj occurs in Ilemicentetes. The 

 talon is still indicated oidy in the case of M. 1 ; the external 

 cusps, with the exception of two, on the anterior and posterior 

 margin, are to all intents and purposes suppressed ; the united 

 para- and mctacone is of considerable height. Moreover the 

 teeth are compressed longitudinally, parallel to the longitu- 

 dinal axis of the body, the last two upper molars less so than 

 the rest. — Geojale is not known to me by personal examina- 

 tion. 



After the above analysis of the molar components in the 

 Centetidae it is not difficult to give an answer to the question 

 with which we started — namely, whether the form of molar 

 that occurs in this family (as also in the Solenodontidee and 

 Chrysochloridge) is to be regarded as primitive, as is pretty 

 generally assumed, or, rather, as the result of a reduction. 

 It is manifest that it is only the richly developed external 

 cusps which are a primitive feature, because, indeed, throughout 

 the mammalian class these cusps present themselves as an 

 archaic phenomenon. I do not venture to go so far as Winge, 

 who regards them as homologous with the three cusps of 

 the Triconodon-\ooi\\. But we found them developed to the 

 fullest extent in Laramie teeth, next to which they are most 

 pronounced in the low groups Polyprotodontia and Insec- 

 tivora ; in other orders in the case of their oldest represen- 

 tatives. 



Undoubtedly of a secondary character, and to be regarded 

 as a retrogression, is, according to what has already been 

 stated, the fusion of the " paracone " and " raetacone " — a 

 phenomenon met with elsewhere among mammals only in 

 premolars and rarely in the M. 3 *. The united para- and 

 * It i3 only in Notoryctes, wliich in other respects also is highly 



