Xo. 488] 



HABITS OF THE SIfRKW 



509 



anise oil, and the specimens taken still had fragments of the nut 



Method of Capturing Food.— On several occasions I witnessed 

 the capture of prey. In the case of the voles and the mice, the 

 attack was essentially the same as described l)y Merriam (1S8(3, 

 pp. 106-168) and Morden (1883, p. 283). The hoii^e mouse, 

 being very agile, was not taken in the open, but only when it eiUered 

 the shrew's burrows. I observed this twice. The clumsy vole, 

 on the other hand, was pursued above ground, cornered in the cage, 

 and caught. In each case the shrew seized the animal's ear in its 

 teeth. After the shrew had been dragged around the cage until its 

 victim was almost exhausted, it quickly loosed its hold on the 

 ear, seized the head in the parietal region, and pierced the skull 

 with its teeth. In two cases the prey w^as dragged part way into 

 a burrow after it had been killed. In the third case it was eaten 

 at once at the surface. The brain and cranium were eaten first, 

 then the neck and shoulders. The skin was closely cleaned and 

 rolled back till the tail was reached. The snout, legs, skin, and 

 tail were left. 



Some difficulty was cxp(M'it'nced in making observations on the 

 eating of snails. When beef ov mice were furnished, snails were 

 not touched. Finally, when all other foods were excluded, snails 

 put ill tho cage of a morning were devoured before the following 



ing by this means that the sluvw would (^at snails at that time of 

 year (early in June). I starved it for a .hiy. then in order to keep 

 it at the surface put it into a caur with sandy soil, and gave it a 

 few snails. The snaiK N\<'re large and their ^lu'lK uere har.l. 

 The shrew put its lower jaw into the aperture in an attem])t to 

 reach the snail. Once its forefoot was thrust in. Failiiiu' to get 

 the snail in this way, it set its teeth across the omcr turn of the 

 shell and tried to break it. This it failed to <io in my pn MMiee, 

 but later the same shells were found broken. It -eeiii-- from these 

 observations that in the ca^t of laru'c >hells, iireakinu' i- a last 



