42 THE ZOOLOGIST. 



stroke. It seemed to me analogous to the preliminary tapping 

 of the Mussels which I have recorded as an endeavour of the 

 Oystercatcher to see whether or not its bill would pass between 

 the valves of a Mussel which was gaping slightly in the presence 

 of surrounding moisture. Careful approaching and watching 

 have, however, failed to reveal any Mussel with its valves 

 separated to an extent likely to admit the bill, excepting, of 

 course, at the hiatus on the ventral border. This being so, 

 a new explanation is required of the tapping. More precise 

 observation of the tapping shows that frequently in attacks on 

 the dorsal border, and more rarely in attacks on the ventral 

 border, the bill does not descend in the mesial plane of the 

 shell, but obliquely inclined towards it, so that the bill strikes 

 unevenly the cleft between the valves, and brings pressure to 

 bear more on one valve than on the other. If the Mussel is 

 relaxed and is taken unawares, that valve is depressed below the 

 level of the other in the region of contact. That is to say, one 

 valve is rotated off its neighbour, and an abnormal gap is formed 

 between the edges of the valves which is sufficiently wide to 

 admit the tip 'of the bill. The bill is then pushed in and moved 

 into the mesial plane of the Mussel which can be opened up on 

 the spot, or detached and dealt with in a more suitable place. 

 The full beauty of this little manoeuvre lies in the fact that 

 entrance can be repeated apparently an indefinite number of 

 times to the same Mussel, provided the bill is withdrawn in the 

 way it entered. This is due to the persistent tendency of the 

 Mussel to contract its valves firmly together in the abnormal, 

 rotated position. This, then, appears to be the correct explana- 

 tion of the tapping. It is an attempt on the part of the Oyster- 

 catcher to take a relaxed Mussel unawares, to rotate the one 

 valve off the other by tapping, and thereby to gain admission to 

 the shell. This conclusion was tested by experiments which 

 added the following information. The valves cannot be rotated 

 when the shells are dry and firmly closed. Rotation is im- 

 possible after the shells are warned by touching them, and also 

 when the shells rest evenly on solid rock. While one shell is 

 being experimented upon, the shells in the vicinity are audibly 

 contracting their valves, and it is necessary to go a foot or two 

 away before proceeding to tap another shell. The stroke must 



