576 THE EYE IN EVOLUTION 



punish it with an electric shock if it attempted to attack a crab asso- 

 ciated with another figure. In such experiments the octopus is emi- 

 nently trainable. The form vision of the animal is surprisingly good. 

 It can distinguish a square of 4 cm. from a square of 2 or 8 cm., between 

 a square and a rectangle of equal area, and between figures of various 

 orientation such as three sides of a square, an L, a vertical or horizontal 

 line, a cross, and so on ; curiously it was found that difficulties were 

 experienced in differentiating oblique lines or a circle from a square. 

 Further, a square of 4 cm. was not confused with a square of 8 cm. at 

 twice the distance, a differentiation which indicates some spatial 

 perception. 



The facility of Octopus in learning to differentiate between horizontal and 

 vertical lines and its relative difficulty in differentiating oblique lines or such 

 figures as a diamond and a triangle, suggested to Sutherland (1957) and Dodwell 

 (1957) that the vertical and horizontal axes have a special status in the dis- 

 crimination of shape. On this basis Sutherland advanced a theory that the 

 output from the visual cells of the octopus was so pi'ojected in the optic lobes 

 as to correspond with a vertical and horizontal system of coordinates ; they 

 would thus correspond with the fundamental coordinates of orientation in space — 

 the vertical depending on gravity and the horizontal aligned to the visual 

 horizon,^ This hypothesis would account for some similar experimental results 

 obtained by Fields (1932) and Lashley (1938) on the sense of discrimination in 

 rats ; and it is also interesting that in man, reference to vertical and horizon- 

 tal components seems to be of primary importance, in association, of course, 

 with other systems of coordinates, in referring a point in the environment to 

 the centre of the visual field. 



Somewhat similar visual reactions can be elicited in the cuttle- 

 fish. Sepia (Sanders and Young, 1940) ; and the perception of move- 

 ment by this mollusc is good with an optimum angular velocity of 

 about 7° per sec. (Boulet, 1954). Indeed, it would seem that Sepia is 

 in some ways more amenable to training than Octopus ; if a prawn is 

 presented as prey and placed behind a transparent glass partition, the 

 former will desist attacking after several attempts while Octopus will 

 persistently swim straight into the screen ; moreover, the cuttlefish 

 will pursue a prawn visually round a corner, while Octopus will give up 

 the hunt unless the invisible prey is reached and can be touched by 

 its exploring tentacles (Sanders and Young, 1940 ; Boycott, 1954) 

 (Fig. 730). It would seem, therefore, that the two species vary con- 

 siderably in their dependence on vision for hunting. It would appear, 

 also, that the former possesses considerable intelligence in that it can 

 pursue its purposes by indirect means and shows some capacity for 

 learning. 



There seems little doubt, however, that these capacities have been 

 exaggerated. Pliny — that prolific purveyor of intriguing inaccuracies — in his 



1 p. 669. 



Sepia 



