1052 UAUOX FllAXCIS NOPCSA ON REVERSIBLE 



Until the present investigation only tlie sliape of each 

 isolated foot-bone was dealt with, now it becomes necessary to 

 consider the whole foot. While each separate foot-bone shows a 

 <lecided reversal of evolution, the whole foot as such shows some- 

 thing else. Although several of the carpal and tarsal bones can 

 be identiOed in all groups of reptiles, nevertheless the number 

 and the relative position of the foot-bones continually change. 

 This is why the foot of an Ichthyosaurian can readily be distin- 

 guished from the foot of a specialised Stegocephalian. In conse- 

 quence of the foot-bones always being differently arranged in the 

 diffei'ent groups of reptiles, evolution seems to be irreversible. 



Thus the foot-bones of reptiles show in a drastic manner how 

 in one point of an organ the evolution can be reversible, but 

 irreversible in another. When such a phenomenon occurs in 

 correlated parts of the body, it is admissible to call the case a 

 mixed one. 



The diagram representing the evolution of the foot-bones of 

 reptiles recalls the one of the evolution of the pelvis : — 



Tarsus and carpus 

 poli/gonal cartilaffitions Polygonal bony Bones ivith 



plates. plates. spherical surfaces. 



Primitive Stegoceplmlians 7 > ( Specialised Stegoceplialians 7 



„ Cotylosaiiriaus ) X „ C'otj'losanrs ) ( ITiglier 



■^ 1. reptiles. 



Primitive marine reptiles-<- 



"~;^Specialised marine reptiles. 



§ (5) 2^he development of the stipraorhital region. 



In nearly all the Stegocephalians the postfrontal and the pre- 

 fi'ontal meet above the orbit and exclude the frontal from this 

 opening. It is only in some highly specialised forms that excep- 

 jbions to this rule can be found. First of all the frontal touches 

 the 01-bit in those gigantic and, as Watson (15) proved, specialised 

 forms, such as Capitosaurus, Mastodonsaurus, and Cyclotosaurus ; 

 secondly, this occurs in those Labyrinthodonts that shoAV a very 

 marked broadening of the skull, such as riagiosaarics (text- 

 lig. 9 (4)); thirdly, this occurs in the aberrant Microsiiurian 

 Diplocaidus (text-fig. 9 (2)). In the less aberrant relatives of 

 Flagiosaitrus and Diplocaulus as, for example, Batrachosuchus 

 (text-fig. 9 (3)), Dlceratosaurus, and Batrachiderpetum (text- 

 tig. 9 (1)), the frontal is yet excluded from the orbit. 



Other Stegocephalians, in which the frontal likewise borders 

 the orbits, are Gephyrostegus, which is characterised by the 

 thinning out of the cranial i-oof, suggesting the formation of 

 temporal vacuities, and Trematojys and Broiliellus, that both recall 

 the Cotylosauria. From all this it becomes evident that in the 

 Stegocephalia the entry of the frontal into the orbits is a sign of 

 fipecialisation, For the sake of convenience one can call the type 



