426 MR. E. S. KUSSELL ON THE RIGHTING 



67 sees. One took the lead, other two attached, one of these gave up, turned 

 on adjacent pair. 

 225 „ Deadlock IV. (see below, p. 429). 



80 „ Two adjacent attached, tips curling in same direction, joined by a 



third, one of original pair gave up, turned on adjacent pair (second 

 choice). 

 80 „ Turned on two adjacent, delayed by one opposite. 



102 „ Turned on one, delayed by an opposite arm holding on. 



120 ,, ,, delayed by an opposite pair holding on. 



82 „ Four attached, turned on adjacent pair. 



Specimen K : ca. 28 mm. 18/6/] 9. 



55 sees. One, three, turned on adjacent pair, hindered by opposite arm. 



45 ,, One, two, turned on adjacent pair. 



25 „ Turned on adjacent pair. 



25 „ One, two, turned on adjacent pair. 



50 „ One, two, three, four, turned on second pair (adjacent). 



42 „ Turned on adjacent pair, impeded a little by opposite pair. 



52 „ First on one, then superseded by opposite pair, on which it turned. 



90 „ At first one arm turned right back, then opposite pair took hold and 



superseded first, which retracted tube- feet. 



60 „ One arm, superseded by opposite pair. Unifying impulse noticeable. 



Specimen L: ca. 13 mm. 18/6/19. 



125 sees. One, two, turned on adjacent pair, long period of incoordination. 



30 „ Turned on adjacent pair. 



55 „ „ „ 



60 „ „ „ impeded by opposite arm. 



30 ,, Turned on adjacent pair. 



83 „ First one attached, then turned on adjacent pair. 

 52 „ 



52 „ 



85 „ Delayed reaction. Turned on two, impeded by opposite arm. 



37 „ Turned on adjacent pair. 



Note. — Specimen J seemed to be affected by the hot sun, so after the first four 

 trials it was kept in the sliade of a stone, as were also specimens K and L. 



From these records it is apparent that while there is great 

 variation in the way the reaction is commenced, there is great 

 uniformity in the way in which the turning is finally accom- 

 plished — almost all turn on an adjacent pair of arms, the actual 

 percentage in the 107 trials recorded being 92-5. In four trials 

 the turning took place on one arm which bent back from the tip 

 progressively : turning took place on three adjacent arms in four 

 cases, including three of the " deadlock " turns, and in two other 

 cases (counted as turnings on an adjacent pair) three arms were 

 engaged until a late stage in the righting, when an outer one 

 gave up. 



When placed on its back Asterina immediately flattens out, 

 then curls the tips of the arms downwards, raising itself a little. 

 The tube-feet are extended and feel about actively. Any number 

 ■of the arms may attach themselves to the bottom by their ter- 

 minal tube-feet. Of these arms two or three generally get a lead 

 over the others. The simplest case is when two adjacent arms 

 get ahead and keep ahead of the rest. Provided the ventral 

 surfaces of these two arms both tnru inwards towards one another 



