CHAP. V. AND DENUDATION. 253 



course impossible to know how much the 

 fragments may have been worn before they 

 were swallowed. It is, however, clear that 

 worms do not habitually select already 

 rounded particles, for sharply angular bits 

 of flint and of other hard rocks were often 

 found in their gizzards or intestines. On 

 three occasions sharp spines from the stems 

 of rose-bushes were thus found. "Worms kept 

 in confinement repeatedly swallowed angular 

 fragments of hard tile, coal, cinders, and even 

 the sharpest fragments of glass. Gallinaceous 

 and struthious birds retain the same stones 

 in their gizzards for a long time, which thus 

 become well rounded ; but this does not 

 appear to be the case with worms, judging 

 from the large number of the fragments of 

 tiles, glass beads, stones, &c., commonly found 

 in their castings and intestines. So that 

 unless the same fragments were to pass re- 

 peatedly through their gizzards, visible signs 

 of attrition in the fragments could hardly be 

 expected, except perhaps in the case of very 

 :soft stones. 



I will now give such evidence of attrition 

 as I have been able to collect. In the 



