460 ADAPTATIONS TO MEDIA AND SUBSTRATES 



from rotating somewhat under the spectacle, and the presence of eye 

 movements associated with a motionless spectacle should never be taken 

 as proof that the spectacle is not of the secondary type. 



Histologically, the secondary and tertiary types are more easily told 

 apart with certainty. It is obvious that if the spectacle of a given fish 

 represents a pair of vertical lids whose aperture has been entirely abol- 

 ished, there can be no connective-tissue strands crossing the space be- 

 tween cornea and spectacle. More important still, there is bound to be 

 a thin epithelium lining this space, covering the cornea (where it repre- 

 sents the corneal epithelium — stratified, of course, in other fishes) and 

 continuing over the inner surface of the spectacle, where it represents the 

 epidermis of the palpebral conjunctiva (see Fig. 15 Id, p. 451). Such a 

 situation has hitherto been reported for only one fish. Some years ago 

 a Dutch investigator, Hein, described and figured an epithelium-lined 

 subspectacular space in a clupeoid, Engraulis sp., and suggested that the 

 same situation might obtain in certain other fishes, specimens of which 

 he was unable to obtain for sectioning (Fig. 152b, p. 454). 



The writer has been unable to see an epithelium on the cornea, or 

 lining the spectacle, in sections of small museum specimens of Engraulis 

 mordax kindly furnished by Dr. Hubbs; but he has found it in the 'pre- 

 herring' Chanos chanos, and is willing to assume that the spectacles of 

 the anchovies and their near relatives {Ancboviella and Engraulis; Etru- 

 meus) are likewise of the tertiary type. The smallest specimens of such 

 fishes fail to show any aperture in the spectacle, which thus either closes 

 before hatching or is intact from its first appearance in the embryo. 



In an acanthopterygian fish (Polydactylus octonemus) , also, the writer 

 has found the crucial epithelium. Here, it apparently consists of two 

 layers of extremely flattened cells — in Chanos, there is but one layer, as 

 in spectacled reptiles. Presumably the spectacles of any fish of the group 

 to which Polydactylus belongs (the percoid family Polynemidae) would 

 reveal the same epithelial lining and completely empty subspectacular 

 space. There is one clupeoid, however, whose condition cannot be so 

 confidently predicted. This is the aberrant Gonorhynchus, which, though 

 related to families which have vertical lids or tertiary spectacles, is 

 catfish-like in habitus and habits. The spectacles of Gonorhynchus may 

 prove to be of the secondary type, rather than the tertiary. 



These teleostean tertiary spectacles have certainly originated from 

 paired, independent, 'adipose' lids which have fused edge-to-edge. Their 

 tissue is of the same sort, and they are likewise devoid of an epidermal 



