574 HIGHER FISHES 



the internal rectus has a more anterior origin (primitive? — cf. lampreys, 

 Chimcera) , migrating backward during growth. The eyeball is somewhat 

 ellipsoidal in Lepisosteus {e.g., 19.5mm. horiz. X 17.5mm. vert. X 15.0 

 mm, axial), but is spherical in Amia. 



The scleral cartilage is hyaline and thick, but is thinned fundally in 

 Amia where it surrounds the 'chorioid gland' (v.i.). The cornea is like- 

 wise thick, as in large-eyed teleosts. The fibrous substantia propria is 

 homogeneous; there is no canal of Schlemm. Descemet's mesothelium 

 thickens at the iris angle to form a massive annular ligament, a cushion 

 of epithelioid cells (said to contain glycogen) applied to the cornea. 

 This thins abruptly to reflect onto the anterior surface of the iris, which 

 thus has a mesothelial facing extending nearly to the pupil margin. Amia 

 is the lowest vertebrate for which the presence of such a layer on the 

 iris can be asserted with any assurance — and it is by no means certainly 

 present in all forms above the holosteans. 



The chorioid of Amia (but not of Lepisosteus) is modified by the 

 presence of a chorioid gland. This structure, which is highly character- 

 istic of the teleosts, is not a gland but rather is a great mass of capillaries, 

 a three-dimensional rete mirabile. Its function is unknown, but it is prob- 

 ably not primarily nutritive. The best guess so far made is that it serves 

 to smooth out the fluctuations of intra-ocular blood pressure which the 

 heart-beat tends to produce, and thus insures a smooth flow of blood in 

 the chorioidal vessels supplied from it, freeing the retina from mechan- 

 ical disturbance. In Amia, as in teleosts, it is shaped like a bloated horse- 

 shoe, straddling the optic nerve with its opening directed ventrally. It is 

 larger in Amia than in any known teleost, and is responsible for the 

 spherical shape of the eyeball of Amia — whose actual intra-ocular cavity 

 is flattened antero-posteriorly as in fishes generally. 



There is an argentea, present only ventro-temporally in Lepisosteus 

 but complete in Amia, where it splits into two layers to enclose the 

 chorioid gland. The innermost of these layers comes almost close enough 

 to the back of the retina to serve as a tapetum (like that in sturgeons) 

 if it were wanted — but since most of the pigment in the chorioid is con- 

 centrated in a thin layer just outside the choriocapillaris, the inner ar- 

 genteal layer is hors de combat as a reflecting device. Besides the chor- 

 ioid gland, the argentea, and the usual vascular, pigmented connective 

 tissue, the Amia chorioid contains many small venous sinuses. 



There is a very narrow ciliary zone between the ora terminalis and the 

 point where the uvea definitely bends away from the sclera to become 



