706 DR. c. I. roBSTTH MAJOR OX THE [June 1, 



old row in some genera, xi/.., Allomi/x^ and Pseiulosciurusi ', the 

 pattern of molars of the last named he considers to be the most 

 primitive amongst known Rodentia '. He places Pseudoscmrus in 

 the Anomaluridae, from wliich family he derives tlie HystricidiB and 

 Dipodidfe, and from the latter the Myoxidse and Mnridse^. 



I have many years ago^, and so hid Hensel " before me, drawn 

 attention to the great resemblance of the molars of Psnudoscmrus 

 to those of Ungulates ; they were compared by me in the first 

 line with the molars of " ffi/racotherium siderolithicum." With 

 regard to AUomys I refer to a more recent paper of mine ". 

 Besides, I wish to draw attention to a fact, which w"ill be more 

 fully considered by me in another place, viz., that amongst the 

 Sciuropteri we equally meet with traces of these ancient outer cusps, 

 namely in S. pearsoni '', and especially in^. xantldpes, Milne-Edw., 

 of which a less worn dentition than that figured by Milne- 

 Edwards'* lies before me (B.M. no. 95. 7. 5. 1). 



These ancient cusps are fm-ther met with in Aplodontia, in 

 whose premolar and molars the middle outer cusp in upper, and the 

 middle inner cusp in lower molars, Winge's 2, are the most con- 

 spicuous of the three. Coues considered Apilodontia to be a very 

 primitive genus, adducing for one of his reasons that the molars 

 are of the most simple type". There is no doubt that this genus 

 is a very low form of Eodentia '", as shown by the skull — in spite 

 of its highly fossorial specialization — and by the structure of the 

 molars ; but not for the reason adduced by Coues ; for in an 

 unworn condition, as figured by Schlosser ", they are shown to be 

 of a complicated type. 



We have next to face the question, what becomes of these 

 ancient outer cusps in the upper molars of Mammals generally ? 

 Years ago I tried to show that the vertical ridges on the outer 

 side of the molars of modern Ungulates are not the unimportant 

 parts which they are generally held to be '^ ; and Winge has since 

 identified them as the homologues of the outer series (1,2,3) of 

 Insectivora and Polyprotodontia ". In proportion as the next 

 .followijig inner cusps, 4 and 5, increase in size and at the same 

 time apparently move outwards, the outer cusps decrease and 

 either become fused with 4 and 5, or persist between them in the 



' ' Gnavere fra Lagoa Santa, etc.,' p. 114. 



■' lb. -p. 116. 3 lb. p. 110 etc. 



* Pahrontograpliica, vol. xxii. 1673, p. 7t>. 



^- Zeitschr. cl. deutsch. geol. Ges. viii. 185B, p. 6fi4. 



" "On Bome Miocene Squirrek etc.," P.Z.S. 1893, pp. 192, 193. 



"^ lb. pi. viii. fig. 20. 



" 'Rech. p. servir i I'Hist. nat. des Mammiferes,' Paris, 18G8-74, p. 171, 

 pi. 15 A. fig. 3. 



^ E. Coues and J. A. Allen, "Monographs of North Amcrifan Rodentia," 

 Report Un. St. Geol. Survey of the Territories. Washington, 1877, p. 555. 



'" See Winge, ' Gnavere fra Lagoa Santa,' pp. 108, 110, 115. 



" M. Schlosser, "Die Nager des europais^ben Tertiiirs," Palaeontographica, 

 Mxi. 1884, p. 106 ',124). 



^■^ " iS'ageriiberreste aus Bolinerzen Siiddeiitscblands und der Schweiz," Pala;- 

 ontograpliica, xxii., 1873. 



" •• Om Pattedyrenes Tandskifte etc.," I.e. pi. iii. 



