'654 



HANDBOOK OF 1'HYSIOLOGY 



M'Tkoi-insioi oc;y hi 



sarj and sufficient condition for having the moon 

 'shrink' at the zenith was elevation of regard, a size- 

 impression conditional upon a particular posture of 

 the eyes in the orbits. Not only direction of gaze, but 



changes in accommodation can condition apparent 

 size [although accommodation is demonstrably not 

 the sole determinant of size constancy (504)]. This 

 need not mean that we integrate some proprioceptive 

 feedback from eye muscles with the visual impression 

 received; what may matter here again is the efferent 

 activity in its influence on the sensory system. 



Consider the micropsiae and macropsiae induced 

 by drugs. If one instills atropine into the eye, thus 

 paralyzing accommodation, the lens is permanently 

 accommodated for distance. If one looks at a land- 

 scape and tries to accommodate (unsuccessfully, be- 

 cause of the paralysis) onto a near object (say, one's 

 finger), the distant landscape shrinks (micropsia). 

 The opposite effect can be obtained by instilling 

 physostigmine into the eye, inducing a spasm of ac- 

 commodation. The lens is then permanently accom- 

 modated for near objects. Attempts at accommodation 

 onto distant objects lead to marked macropsia; the 

 distant landscape expands. These forms of micropsia 

 and macropsia can perhaps be understood if one 

 considers what happens as long as accommodation is 

 normal. As objects approach, two continuous and 

 concomitant changes occur; accommodation in- 

 creases, and so does the visual angle subtended by 

 the object. Perceptually, however, the object stays 

 approximately the same size. One can assume, as we 

 have several times before, that this constancy is the 

 result of a central counteraction which 'shrinks' the 

 approaching object in keeping with increasing ac- 

 commodation. Such regulatory feedback need not go 

 via the periphery; a corollary discharge from the 

 central oculomotor system into the visual system 

 could be conceived .is the source of such a 'physiologic 

 micropsia' for approaching objects, and of the relative 

 macropsia for receding ones. The notion which has 

 alreadv been developed in some detail by von Hoist 



,. . 1 1 is schema ti/ed in figure 31 . The same principles 



have been used in Mittelstaedt's analysis (348) ol pre) 

 capture in mantids. 



h is clear thai on these notions many illusions 

 mil .is micropsia and macropsia) are really con- 

 stancies misapplied ' < >nc of the simpler (and earlier) 



" Analogous jiderations apply to nunc- complex forma of 



illusions such .is the size-weight illusion (75, 76). In tins 

 familiar illusion, .1 large (but liollnu 1 object is judged .is mui li 

 lighta (upon hefting than it would be it the identical weight 

 were presented in the form ol .1 less l>ul'\ (but more massive) 



examples for this relation is Emmert's "law" (52, 112) 

 which states that an afterimage increases with the 

 (perceived) distance of the surface on which the 

 observer projects it. If our theoretical approach to 

 size constancy is valid, then the increasing apparent 

 size of afterimages, at increasing projection distances, 

 is such an instance of size constancy running off 

 in vacuo. The 'retinal size' of the afterimage does not 

 change, once established; it is just for that reason 

 that the perceived size of the afterimage is subjected 

 to the (centrally guided) expansion with increasing 

 distance. For a normal object, this expansion in size 

 would counteract the diminution in visual angle as 

 the object recedes. Convergence and accommodation 

 are among the principal determinants here of after- 

 image size. Some observers, however, report the 

 same effect while projecting an afterimage into in- 

 creasingly 'distant' recesses on the perspective 

 drawing of a tunnel. 



Muslim*.: Phenomena and Interpretations 



ILLUSIONS AS MISAPPLIED CONSTANCY EFFECTS. The 



famous optic-geometric illusions, as we have seen 

 (above, p. 1601), have been a major concern of the 

 nineteenth-century students of perception (53), the 

 interest waning somewhat (though never completely) 

 with the insight into the inappropriatcness of sepa- 

 rating these illusions from other perceptual phe- 

 nomena. Two major types of illusions were dis- 

 tinguished, those involving visual extents, and those 

 involving angles, although manv patterns were 

 devised incorporating both angular and linear dis- 

 tortions in one and the same figure. One of the 

 earlier illusions of extent was Oppel's demonstration 

 (368) in [855 thai interrupted lines are overestimated 

 in comparison with uninterrupted ones dig. 34, 

 upper left). Soon afterwards in 1858, Wundt (550) 

 pointed out that vertical lines tend to be overesti- 

 mated in comparison with horizontal lines, is shown 

 in figure 34. von Helmholtz (501) combined both 

 effects that of vertical-horizontal, and of inter- 

 object Presumably we 'expect' objects to show increasing 

 weights with increasing size; in holding out .1 hand to receive a 

 weight or to lift it, we are 'set' in different fashions, depending 

 on tiie anticipated weight. Unless then- are unanticipated 

 disproportions between the actual densitj and the visual 

 appearance of .1 weight, our motor readiness will be appro- 

 priate In fact, there are marked (though relatively unexplored) 

 constancies for weights, e g we ordinarily compensate, in part, 

 loi differences in lever-action, as when the same weight is 

 applied to more proximal or more distal parts ol a limb (124, 

 250). 



