PETERSON : DESCRIPTION OF NEW RODENTS 187 
of the compact mass of roots, which are to be spoken of later. In two cases skulls 
were found near the rounded ends of ‘rhizomes,’ and crowded close to the inside 
surface of the surrounding silicified zone of roots. The type-specimen of Stencofiber 
fossor (deseribed in the preceding pages) was also found near the end of one of those 
“rhizomes.” 
The accompanying figure (Fig. 6) is a diagrammatic illustration intended to repre- 
sent an actual specimen of an apparently incomplete Demonelix in the Carnegie 
Museum collections. ‘This specimen was found by the writer, and presents some inter- 
esting features. It is a very short spiral, much weathered, and contains part of the 
skeleton of Stencofiber fossor. This short spiral was found, in situ, between two layers 
of more coherent material than that in which it was imbedded. The thin stratum 
below the lower end of the spiral (Fig. 6, b) is a horizontal layer of organic structure, 
some three or four inches thicky and intermixed with sand. This organic structure 
resembles some sheets of vegetable growth covered by a sudden influxes of sand 
brought in by water, and was simply a portion of the deposit in which the burrow 
was afterward made. The stratum overlying this specimen (Fig. 6, a) is also quite 
horizontal, but consists of much harder, silicified material, and is more irregular in 
thickness. Such hard layers of sandstones are 
quite common in these beds. The upper end of a 
this specimen of Demonelix gradually loses its 
character, and cannot be traced upward into this 
hard cap of sandstone. The latter was probably 
formed by the deposition of sand in a pool of a 
water on the surface of the ground. The exit 
of the Demonelix hole naturally lost its shape 
ie b Cc 
under these conditions. =O) 
The more perfect specimens of Deamonelia Ns 
Forms of Demoneliax sometimes 
possess so-called “rhizomes,” which are nearly found. Taken from field sketch of exposed 
always larger in circumference than the ver- ‘pecimens lying on the surface with only the 
Fi 6 ; 3 ; Pe base of the vertical spiral attached to the 
tical spiral. Occasionally, this ‘rootstalk”” is 
“rhizome.’’ a to 6, 18 inches ; b toc, 14 feet. 
branched, two, three, and even more times. Each 
of the branches is of approximately the same general size, with local enlargements 
and pockets up to the extreme blunt and rounded end. No instance was found of 
these “rhizomes” crossing one another, so as to form a short kink, such as is often 
seen in roots of recent plants. ‘The end of the rhizomes is often found enlarged 
in a more or less hemispherical manner. I have often traced them for six, and 
sometimes for fifteen feet from the base of the spiral to their end. 
