580 HIGHER FISHES 



A further major complication is introduced by the common or usual 

 presence of an 'autochthonous' layer or mass of coarse-fibered sub- 

 stantia propria. This is distinct from and internal to the scleral layer, 

 and always itself bears the thin Descemet's membrane and mesothelium 

 (lacking in pikes?) . Rarely, the scleral layer may be easily separable from 

 the autochthonous, and may even (e.g., Lepomis macrochirus) bear in- 

 wardly a mesothelium for all the world like a Descemet's, though with- 

 out any elastic, cuticular basement membrane. The phylogenetic origin 

 of the autochthonous layer cannot be traced. It looks as if it had been 

 formed in situ from 'nothing' (hence its name), magically interpolated 

 between Descemet's membrane and the scleral propria. At the periphery 

 of the cornea, the autochthonous layer usually thickens greatly, then 

 abruptly tapers to a knife-edge termination opposite the front margin of 

 the scleral cartilage or bone. Descemet's layers ordinarily do not extend 

 nearly this far peripherally, for the mesothelium is reflected over (or 

 forms) the annular ligament (see also p. 574) and continues back toward 

 the pupil on the anterior face of the iris. 



In a few fishes, including gobies and particularly the plectognaths 

 (trunkfishes, puffers, ocean sunfishes etc.) the corneal substantia propria 

 exhibits a complex lamination, with histologically peculiar intercalated 

 layers which cannot be related at present to the typical lamination- 

 system just described. 



The 'annular ligament' — an inappropriate name, but one for which 

 no good substitute has yet been offered — is almost universally present; 

 but it can be greatly reduced or lacking in species of a genus which 

 characteristically has it well developed (e.g., Anabas). It is no teleostean 

 monopoly, but was invented by fishes as archaic as the Chondrostei, if 

 not by the cyclostomes. The justification for calling it a ligament is the 

 fact that it gives the appearance of forming a bracket between cornea 

 and iris, holding them at a fixed angle to each other. Actually, the tissue 

 of the ligament is (always?) so very ductile that it can have no such 

 sustentative function. The ligament adds to the difficulty of defining 

 the boundary between the iris and the ciliary body in teleosts, for one 

 naturally tends to consider, as 'iris', only what is free of the annular 

 ligament. Actually, the greater part of the iris — best defined as the por- 

 tion of the uvea within the limbal circle — is covered by the ligament, and 

 the true ciliary zone (from limbus to ora terminalis) is very narrow. 

 Embryologically, the annular ligament arises from an accumulation of 

 mesodermal cells which lie at the periphery of, and continuous with, the 



