oo8 GEORGINA SWEET. 



being indistinguishable. In two eyes a small oval nodule of 

 liyaline cartilage (fig. 4, c, ii), containing some six cells, is 

 to be found lying in the sclero-choroid, just on the inner 

 side of what ought to be the exit of the optic nerve. In 

 another case a similar though smaller nodule was found lying 

 in front of the eye, a little to its outer side. The structure 

 of this sclero- choroid is very much like that of Bhineura 

 [4, p. 537j among others. As in that also this layer is pro- 

 longed proximally, forming a sheath which must represent 

 the covering of the optic nerve, being connected posteriorly 

 to the skull along with the conical capsule. Small bars of 

 cartilage comparable to the nodule are also found in Ehineura 

 and Amblyopsis [1, p. 563]. 



Pigment Epithelium. 



Immediately within this sclero-choroid lies the conspicuous 

 hollow ball of pigment, dense and thick-walled anteriorly 

 and exteriorly, as well as above and below, and very thin on 

 its inner side and posterior end. At first, in the absence of 

 embryonic material, one is led to consider this as a much 

 thickened choroid, but, on comparison with other forms in 

 wliich developmental changes can be followed, it is seen to 

 be probably the greatly changed pigment epithelium of the 

 retina, though there is nothing in the irregular, broken, and 

 jumbled masses of granular pigment to suggest such a well- 

 defined cell-layer as the retinal pigment. No trace of cell- 

 fctructure can be seen enclosing the pigment, though, in 

 occasional gaps between the masses, small oval cells can be 

 seen without any granules within them. There are never 

 any processes inwards or outwards. It will be noted that, 

 as in Troglichthys [1, p. 581], and Typlilomolge [2, p. 53, 

 and pi. o, figs. 1, G, and 7] the pigment is much denser in 

 front, where it might be ex])ected to be absent, and very thin 

 or wanting posteriorly, where it ought to be thicker. In 

 this these three forms differ from all other known degene- 

 rate vertebrate eyes. 'J'he irideal region is much reduced 



