506 



THE AMKRH 'AX SATURALIST [Vol. XLI 



the burrows. Sometimes they were thrust into the soil at the sides 

 of the burrow in groups of 2 to 10. At other places short branches 

 led either downward or laterally from the main burrows, and then 

 expanded into chambers filled with shells, mingled with loose soil. 

 Such chambers contained in some cases as many as 80 shells. 

 Generally, all the shells were em])ty, l)ut one such chamber con- 

 tained m shells, of which :A were still occupied the last week in 



beneath ti.e main burrow, the branch burrow leading to it was 

 spiral in form, like a winding staircase. ( )ne of these is illustrated 

 in Fig. 2. In two instances empty shells, broken exactly like those 

 found elsewhere, were found iir-^ide nests which, from tlu'ir com- 

 position of slu-(Ml(K-d material and position at the surface of the 

 ground, must have been Mi.rotus nests. 



Method of Burrowing. — The method of burrowing was observed 

 and experimented on in the laboratory, where a shrew captured 

 by hand was kept for some time. When it was first confined, 

 loose bhick soil was place.l in the cage to a depth of about 10 cm. 

 Into this soft soil the shrew at once thrust its nose, and by violent 



through the >oil like a wedue. Xo difficulty was experienced in 

 burrowing 20 or ;',() cm. in a minute. 'Vhv UKnements underground 

 wen- evi.h-nce.l bv the moveuH-nt of the soil at the surface; but 

 no ridge was forined above the burrow. When clods were en- 

 countered, they were rea.lily moved, even if fifteen or twenty times 



opening to the surfa;-e at seven or eight points was worked out. 

 The aggregate length <.f burrow was not known, as 1 did not dig 

 out the soil, l)Ut it include<l the whole cage whirh measured ;;.-> l.y 

 4S cm. 



the end of a week there was an aggregate of 35 cm. of Imrrow 



