282 HOW DO STARFISHES OPEN OYSTERS? 



apparatus had to be so devised that the bivalve, whilst remaining in 

 water, could be placed in such a position that a measurement of the 

 extent of opening of the shell could easily be made. A glass dish 

 containing sea water (g) was placed on a small table. In this one 

 stood a smaller but higher dish, also containing sea water, which could 

 be renewed from time to time from the larger one. A Venus was now 

 surprised, the handle of a scalpel being placed between the two shells 

 before it had time to close them. The bivalve of course, as soon as 

 the scalpel was put in, closed immediately, and held it fast, and so 

 could easily be taken out. Two tlesh-hooks were then taken : a short 

 one made entirely of metal, with two teeth at each end (/), and a 

 second one with a scalpel handle (/'). The two teeth at one end of 

 the short hook were placed in the shell-opening. The teeth of the 

 second hook were also placed in the opening in such a way that they 

 came between the teeth of the short hook. A double string was then 

 slung from the two hinder teeth of the short hook, and a similar string 

 was made fast to the larger one at the beginning of the flat handle. 

 The handle of this hook was placed between the two strings on its own 

 side, so that it was kept in position by them. By this arrangement, 

 with the aid of the friction of the double strings on the edges of the 

 dish (at a and a), the bivalve was held quite motionless, with the 

 opening upwards. The molluscs were not disturbed by this experi- 

 mental strain ; but, as soon as the shell was far enough open, protruded 

 their siphons and commenced breathing, so that one may say that they 

 were simply normal. Attached to the double strings (at d and d') was 

 a single common string, from the middle of which a scale-pan (w) 

 hung. The weights on this scale-pan were, of course, not all effective, 

 on account of the considerable friction at the points a, h, c, and a', b', c'. 

 In order to determine the true effective weight, I afterwards replaced 

 the bivalve by a spring balance, which was pulled out by the weights. 

 Such spring balances never weigh quite correctly ; but in this case 

 one or two grams does not matter, and I give, therefore, in the following 

 tables, only round numbers (friction being allowed for) : — 



1. Venus 4 cm. long, 3-4 cm. broad. 



7.55 a.m., loaded with 900 grams. 



8.10 „ commenced to open. 



8.15 „ open 2 mm. 



8.30 „ open 3-5 mm. 



1.0 p.m., open 3-5 mm. 



G.IO „ open 4 mm. : then set free. 



