Eye 



403 



nents of the chick optic cup by isolating them 

 in vitro. 



Dragomirow, invoking Child's concept of 

 metabolic gradients, believes that the diver- 

 gent development of the two parts of the 

 vesicle is an expression of simple quantita- 

 tive differences in physiological activity im- 

 posed by environmental inflviences. He 

 suggests that in normal development it may 

 be contact with the lens or lens ectoderm 

 that initiates the differential in question, a 

 proposal in keeping with the fact that the 

 distal (retinal) wall of the vesicle is the 

 first to become channeled in a fixed course 

 of development. The polarizing action of the 

 lens is not specific in nature, however, since 

 Dragomirow has shown that the epithelium 

 of the ear vesicle may serve in the same ca- 

 pacity. When an eye cup is grafted adjacent 

 to the ear placode, the portion of the pigment 

 layer in contact with the wall of the develop- 

 ing labyrinth is subsequently induced to 

 form a secondary retina. In fact, Ikeda ('37a) 

 reports that the nasal placode, and perhaps 

 other epithelia as well (e.g., peritoneum), 

 are capable of evoking retina formation 

 through contact with the pigment layer of 

 the eye. 



Holtfreter ('39) has shown that intimate 

 association with mesenchyme is another con- 

 dition that must be fulfilled if the eye vesicle 

 is to realize its potentialities for normal 

 organotypic and histotypic differentiation. 

 Eyes evaginating from medullary plates iso- 

 lated in vitro, under conditions that preclude 

 access to the matrix of mesenchyme that 

 normally surrounds the developing vesicles, 

 attain only the most rudimentary level of 

 organization. His experiments show further 

 that when circumstances permit, the eye 

 vesicle seems actually to "seek" association 

 with mesenchyme, as if in response to a posi- 

 tive physical attraction or affinity between 

 the two. These experiments of Holtfreter 

 serve to expose, perhaps better than any 

 others, the fxtreme limitations of the eye 

 rudiment as an autonomous developmental 

 unit. These limitations are not fully revealed 

 by ordinary tests, such as heterotopic trans- 

 plantation, which leave the vesicle in its 

 normal context insofar as association with 

 mesenchyme, epidermis or other epithelia, 

 etc., is concerned. When stripped of these 

 relationships, the anlage appears to be quite 

 incapable of realizing the potentialities as- 

 signed to it at the time of its first delinea- 

 tion. 



Eakin ('47) has demonstrated that long 

 after the inner layer of the cup becomes 



channeled as sensory tissue, it remains labile 

 with respect to another fundamental feature 

 of its organization. By reversing surgically 

 the outer and inner faces of the sensory 

 layer, Eakin found that the mediolateral 

 axis of its development does not become 

 fixed until the actual beginning of histologi- 

 cal differentiation. Retinae reversed prior to 

 this time develop with sensory and neural 

 layers in proper polar adjustment to the rest 

 of the eye, while those reversed at later stages 

 preserve their original or prospective polar 

 constitution, with rods and cones facing 

 toward the lens. Eakin suggests that physical 

 and perhaps chemical influences exerted by 

 the lens and the pigment epithelium are the 

 possible morphogenetic forces responsible for 

 establishment of polarity within the inner 

 layer of the cup. 



Another aspect of retinal determination, 

 the localization of function in the quadrants 

 of the retina, has been studied by Stone ('48) 

 in experiments involving the 180 degree 

 rotation of the optic cup and lens. When 

 this operation is performed prior to the stage 

 when the embryos first become motile, the 

 larvae subsequently show normal motor re- 

 sponses to visual tests. However, rotation at 

 a somewhat later stage, when the beating 

 of the heart first becomes prominent, re- 

 sults in reversed visuo-motor behavior. Ef- 

 forts to capture moving prey are misdirected, 

 and movements in response to rotating black 

 and white striped drums are the opposite of 

 those shown by normal animals. 



The choroid fissure is another feature of 

 retinal topography that is not yet imprinted 

 or localized in beginning stages of eye de- 

 velopment. This is shown by the failure of 

 fissures to develop in eyes grafted embryoni- 

 cally to heterotopic positions, and by the 

 "host-wise" formation of the fissure in the 

 normal ventral position after 180 degree 

 rotation of the optic vesicle (Sato, '33a) 

 during early tail-bud stages. The vesicle 

 becomes less adaptive in this respect as 

 development proceeds, but occasionally eyes 

 with normally situated choroid fissures de- 

 velop following rotation as late as the stage 

 when contractile responses are first exhib- 

 ited by the embryo. Sato believes that 

 the formation of the fissure at the ventral 

 margin of the cup is determined by the 

 attachment at this point between cup and 

 brain, possibly through exertion of growth- 

 inhibiting influences or even simple mechani- 

 cal traction by the optic stalk. The results 

 of Sato on the determination of the choroid 

 fissure have been confirmed by Barden ('42, 



