568 HIGHER FISHES 



and autonomous and sluggish in their action. This musculature is scant 

 in the deep-sea elasmobranchs, whose wide pupils are almost permanently 

 open. At the root of the iris, the organ makes a slender angle with the 

 cornea. There are no pectinate or annular ligaments in the iris-angle 

 region, no loose meshwork tissue, and there is no canal of Schlemm. 



The Retina — The elasmobranch retina is characteristically pure-rod, 

 with a high ratio of visual to ganglion cells. This great summation helps 

 to confer the photic sensitivity upon which the light-shunning habits of 

 these fishes are based; but it necessitates a low visual acuity — estimated 

 by Franz to be, on the average, 5% of that of man. Franz determined 

 the number of fundal rods to be 10,800/sq. mm. in Raja batis, 21,600 in 

 Torpedo, 24,000 - 75,000 in various small sharks, 100,000 in Chimara, 

 and 132,000 in Etmopterus. Corresponding ganglion-cell estimates were : 

 Ra]a batis, 1500; Torpedo, 5000; the small sharks, 1200 - 3600; Chi- 

 mcera, 600; and Etmopterus, 900. 



The pigment epithelium is devoid of pigment (and usually, of cell- 

 processes) over the whole extent of the sensory retina in all elasmo- 

 branchs except those few which lack the tapetum lucidum for obvious 

 reasons : Lcemargus (an abyssal shark) , Selache (the basking shark) , and 

 Myliobatis (a pelagic ray with a cone-rich retina). 



The horizontal cells in elasmobranchs are massive, much like those 

 of lampreys (see Fig. 162a, b; p. 559). Though fine processes have been 

 seen on them in Golgi preparations, it is unlikely that they have any- 

 thing but a supporting function. The bipolar, amacrine, and ganglion 

 cells are not confined to their 'proper' nuclear layers, but may occur out 

 of position, in layers above or below. Such misplacements are quite char- 

 acteristic of crude, scotopically-adapted retinas in lower vertebrates. 



Cones are known to occur only in one dogfish genus iMustelus), the 

 eagle ray Myliobatis aquila, and (doubtful!) the monk-fish, Squatina. 

 They are least distinct from the rods, morphologically, in Mustelus 

 canis (Fig. 163c, p. 561), better differentiated in M. mustelus, and are 

 completely cone-like {i.e., short, with plump inner segments and small 

 outer ones) in Myliobatis, according to Mile. Verrier. It seems clear 

 that these few modern-elasmobranch cones are 'new', secondary deriv- 

 atives of rods; but, the 'original' vertebrate cone must have persisted 

 through the ancient elasmobranchs, in order to be handed on to the 

 higher fishes — none of which, of course, were derived from sharks, rays, 

 or chimaeras (see Plate I). 



