590 HIGHER FISHES 



sirenidae. Its relatively large eye may have, in particular, a mechanism 

 of accommodation; and its cone oil-droplets may be colored in life. But 

 the animal is reputedly nocturnal (in captivity, at least), and may not 

 have retained such things even though some diurnal ancestor may have 

 had them. Neoceratodus (and Latimeria) remain our chief hope of ever 

 learning the origin of the amphibian mechanism of accommodation, 

 which is so distinct from those of all known fishes. Unless otherwise 

 noted, the following statements apply only to Protopterus icethiopicus) : 



The ca. 2.0mm. eyeball turns freely under a transparent dermal 'sec- 

 ondary spectacle' (Lepidosiren also) . The cup-like scleral cartilage, which 

 is about two cells thick, reaches only to the equator of the eyeball; but a 

 fibrous continuation of it becomes, anteriorly, the inner portion of the 

 cornea. The fibrous layer of the sclera external to the cartilage also 

 continues forward as a portion of the corneal substantia propria, entirely 

 unconnected with the skin of the spectacle (and apparently separate, or 

 at least very readily separable, from the inner moiety of the cornea). 

 The Descemet's membrane and mesothelium are the thinnest imaginable. 



The chorioid consists of little more than a choriocapillaris, with only 

 wisps of connective tissue, containing a very occasional pigment cell — 

 altogether the thinnest, simplest chorioid outside of the blind vertebrates. 

 There are no traces of a chorioid gland or of an argentea. The circu- 

 latory pattern of the eye includes a set of vitreal vessels (not in Neo- 

 ceratodus — hence there, perhaps, a falciform process?) . 



The iris departs directly from the ora terminalis without the inter- 

 calation of any zone which could be called ciliary, and without support 

 for its root in the form of any pectinate ligament or mass of meshwork 

 tissue in the angle between it and the cornea. It is very thin — its stroma 

 thinner than its retinal layers. Even the latter appear to have tried to 

 thin out, for the pigmented anterior layer is squamous rather than 

 cuboidal as usual. The posterior retinal layer is nearly free of pigment, 

 so that the iris (and indeed, the whole eye) is as simple as that of a 

 brook lamprey. The relatively huge (1.16mm.) lens lies entirely behind 

 the iris, so that the pupil is free to change in size; and it can do so, in 

 Protopterus at least, despite the total absence of any discernible mod- 

 ification of iridic cells into myoepithelial elements. 



The Dipnoan Retina — Here again, little can be said about Neocer- 

 atodus, and not much more about Lepidosiren. In the latter and in Pro- 

 topterus, the pigment-epithelial cells are huge, the epithelium being as 

 thick as the sclera and much thicker than the rudimented chorioid. 



