PETERSON : DESCRIPTION OF NEW RODENTS 189 
It seems reasonable, that, after our Miocene rodents vacated their burrows, the 
roots of plants growing on the surface would find their way down the spirals as well 
as down the straight shafts. The tamped walls remained harder than the surround- 
ing ground, and when the roots and rootlets reached these walls, they followed the 
line of contact between the walls and the inside filling, perhaps similar to the roots 
C 
a 
Fia. 9. A newly dug prairie dog burrow. From a field diagram. c, Exit of burrow; a, nearly horizontal 
tunnel about 5 inches in diameter ; d, nest 8 inches in diameter and four feet under surface ; }, cross-section near nest 
showing emargination of cast caused by bedding. From c to d, 9 feet. 
of a potted plant. Occasionally a root would pierce the wall, and continue on its 
way to the next coil. This is seen on nearly all specimens of Demoneliz. As has 
been stated before, there is a cylinder of tangled roots on the outside surface of 
Demonelix. Toward the center of the casts the roots are much less numerous. 
a 
Fia. 10. Fria. 11, 
Fia. 10. Diagram of cast of an old burrow of Prairie Dog. a, exit; b, at this place the burrow was caved in. 
From a to b, 10 feet. 
Fic. 11. An underground fortress of a mole, Zalpa europxa; after J.G. Woods, in ‘‘Homes Without Hands.’’ 
n, nest. 
In speaking of the recent mole, J. G. Woods (in ‘“‘ Homes Without Hands”) says 
that ‘we do not generally know the extent or variety of this animal’s tunnels or 
that it works on a regular system, and does not burrow here and there at random.” 
In this volume is an illustration of a mole-hill (Fig. 11), which is represented (by per- 
