18 RE VISION OF AMER ICA N MOLES— TR TIE. 



Some considerations of a different character may throw light on the 

 problem as far as concerns Scalops. It will be noticed upon examina- 

 tion of forestry maps of North America ' that the coniferous forest ter- 

 minates on the south at the exact point where the moles reach their 

 northern limit. It would appear as if these forests prevented the 

 further northward distribution of Sca]o2>s. This is not at all improb- 

 able, though it is likely that their effect is indirect rather than direct. 

 Adopting this view, we may look upon ^Scalops as a northern genus 

 whose proper northern extension has been reduced by the growth of 

 coniferous forests.'* As the whole family Talpidw is of northern origin 

 so far as known, this view is perhaps not without force. 



Family TALPID^. 



Diagnosis. — Small insectivora. Humerus articulating with both 

 scapula and clavicle. Clavicles redm-ed in length. An elongated 

 manubrium sterui. Acetabula closely approximated. Eyes minute. 

 No ear-conch (except in Uropsilus). Molars three on each side of either 

 jaw; incisors not exceeding four; premolars not exceeding four. Molars 

 broad, with W-shaped external cusps, and an internal basal ledge. 



Zygomatic arches complete. Tympanic bulhie more or less complete. 

 Lumbar vertebrii?" with hypapophysial ossicles. No symphysis pubis. 



The first paragraph of this diagnosis contains all tlie characters by 

 which this family can be clearly separated from the other families of 

 the order. The typical moles, such as Talpa, Scalops, etc., are, of 

 course, readily separable, from the representatives of other families, 

 but many of the distinctions break down in such forms as Xeiirotrichus, 

 Uropsilus, etc. 



Thus, the radial sesamoid, characteristic of the manus of the typical 

 moles, is absent in Uroirichus, Uropsilus, etc. An ear-conch, which is 

 absent in the typical moles, is present in Uropsilus. 



INIany characters which are sufficiently diagnostic in combination 

 are, nevertheless, shared by the representatives of other families. 

 Thus, the peculiar hypapophysial ossicles of the lumbar vertebra? occur 

 also in the Erinmeidie? Complete zygomatic arches occur in Chrijso- 

 chloridfv and Erinaceidfr as well as in the moles. The CJirysochloridw 

 and Potamogalidiv, like the moles, are without true symphysis pubis. 



The moles are on the whole more nearly related to tlie shrews than 

 to any other insectivores, and the more aberrant forms, like Urotrichus 

 and Uropsilus, bear a strong external resemblance to the shrews. Sjtill 

 there are no moles without zygomatic arches and none with the long- 



'See Forestry Report, Tenth Census, by Sarf;eut. 



" It is a sinunlar fact that where the coniferous trees are cut down, as, for example, 

 in Alaska, they are replaced by deciduous trees. When the woodman's ax has passed 

 over the Alaskan forests, therefore, we may expect to lind that territory covered by 

 deciduous trees. 



•'What is more remarkable these ossicles occur also in rerameles, a marsupial mam- 

 mal. 



