258 Bulletin American Museum of Natural History. [\o\. XXVII, 



ences which are eciually striking and more significant of the wide genetic 

 separation in the two forms. For example, in the humerns of both forms 

 the entocondylar process is greatly produced. But this is only an exaggera- 

 tion of a feature that is common among fossorial mammals, {e. g., Orijcter- 

 opus, Dasypus, Spalax) and is associated ^\■ith very striking differences in 

 the other parts of the bone {e. g., in the deltoid and supinator crests, greater 

 and lesser tuberosities). The long, hook-like olecranon of the ulna is efjually 

 a fossorial character which is realized to a less extent in Spalax and Dasypus. 

 The scapida of Noforyctes is of the broad fossorial type, with a second or 

 posterior spine analogous to that in Edentates, while the scapula in Chryso- 

 chloris is of the Sorex type with high spine, narrow blade and broad l)ifid 

 acromion. The j)elvis of N otoryctes approaches that of Chrysochloris in the 

 great reduction of the thyroid fenestra, yet it retains strong evidence of 

 derivation from the Marsupial type, especially in its vestigial marsupial 

 bones (Stirling, LSOl, p. 17S), while the pelvis of Chrysochloris, presents 

 no feature inconsistent with derivation from the Microgale type. The pes of 

 N otoryctes and Chrysochloris are very unlike, the pes of the former showing 

 some traces of the Marsupial syndactyly (Winge, Dollo; Bensley, 1903 p. 

 172) while the pes of Chrysochloris conforms to the Insectivore type. 



These resemblances and differences have been cited at some length, 

 partly in illustration of the principles discussed on pages 107, 111, 463. 



The occurrence of Chrysochlorids in the Oligocene of North America 

 has been recorded by INIatthew (190G) who referred to that family the fol- 

 lowing forms: Xcnotherium unicum Douglass (1900), based on a skull 

 without teeth; Apternodus meclioevus INIatthew (1903.1) based on a fragment 

 of a lower jaw with several molar teeth, Microptcrnodus horealis Matthew 

 (1903.1) an imperfect lower jaw; Arctoryctcs ierrenus ^Matthew (Lower 

 INIiocene) based on a humerus \\ith marked Chrysochlorid characteristics. 

 The skull of Xcnotherium (cf. Douglass, 1906, ])1. xxi) resembles that of 

 Chrysochloris aurea (cf., Dobson, 1883, pi. xi) not only in general appearance 

 but especially in many significant details: such as the hemispherical form of 

 the bullje, the form of the snout, zygomatic arch, peculiar lateral occipital 

 crest on the squamosal, etc. The only im})ortant dift'erence is in the char- 

 acter of the cheek-teeth alveoli, which in Xcnotheriuvi are circular, in Chry- 

 sochloris triangular. On the other hand Xenotheriinn parallels the Miocene 

 Talpid Proscalops [cf. p. 265) in its brachycephalic proportions and in several 

 of the characters noted above, but the molars must have been very different. 



In regard to Apternodus, Matthew now (1909) describes it as having 

 "Molars as in Ccntetes, premolars reduced to three." The possibiHty that 

 the genus may be a member of the Centetoidea is also indicated by the follow- 

 ing observations: 



