MEDIAN EYES IN EXTINCT VERTEBRATES 335 



nerve connecting it with the central nervous system (Fig. 134, Chap. 17, 

 p. 188, and Figs. 183, 185, Chap. 20, pp. 259, 261), and (b) those in which 

 the parietal sense-organ was constricted off from its stalk by the growth of 

 the skull and was left as a vestigial cyst (or pair of such cysts) in the sub- 

 epidermal extracranial areolar tissue (Figs. 162, 167). In the latter all 

 connection with the central nervous system would have been severed and 

 the parietal vesicle would have been quite functionless as a sensory organ. 

 In some instances an impression or pair of impressions was left on the dorsal 

 aspect of the pineal plate (Fig. 229) ; in others the development of the 

 vesicles may have been less pronounced, and though they may have 

 persisted in the adult animal — as is the case with Stieda's " frontal organ " 

 in the frog — they left no impression on the outer surface of the skull. 



On comparing the pineal pits and impressions which have been 

 observed in the skulls of fossil Ostracoderm fishes with those in the 

 skulls of living species, it may be concluded that certain of the more 

 ancient extinct specimens fully confirm the evidence which has been 

 independently obtained from both embryonic and adult specimens of 

 the bilateral origin of the pineal body. The palaeontological evidence 

 also points to regressive changes having already commenced in some of 

 the most primitive and earliest known fishes, and having continued in 

 their descendants until the present time. Nevertheless vestiges of the 

 pineal system have persisted in nearly all types of living vertebrates, and 

 structural changes have taken place in the organ which have suggested 

 the occurrence of an evolutionary transformation into an organ of internal 

 secretion. These changes are, however, much more pronounced in the 

 embryonic stages of development and the early post-natal period of life 

 than they are in the adult animal, in which evidence of regression is nearly 

 always present. 



Now, it is interesting in connection with the general occurrence of 

 regressive changes which have taken place in the pineal system since 

 the period when it appears to have attained its highest degree of develop- 

 ment, in the extinct amphibians and reptiles of the late Palaeozoic and the 

 Mesozoic periods, to recall that among the existing Dipnoan fishes — 

 Lepidosiren, Protopterus, and Ceratodus — which in some respects appear 

 to be intermediate between fishes and amphibians, their general features 

 when compared with the early fossil fishes have remained essentially 

 unchanged since the middle of the Devonian period, and such changes 

 as have been observed are regressive rather than evolutionary. According 

 to the description given of them by Smith- Woodward, " they have in 

 the interval merely abandoned the fusiform shape which is adapted for 

 a free-swimming life and become more or less eel-shaped in adaptation 

 to a wriggling and grovelling existence at the bottom of the rivers into 



