248 Mr. G. A. Boulenger on the 



in my opinion, and as I expected from the author's preliminary 

 publications, failed in its object. 



I will merely allude again to our differences of opinion as 

 to the probable derivation of the various forms dealt with. 

 I have already expressed my views on this matter *, and I 

 see no reason to alter them. I adhere to the general lines 

 first laid down by Eimer, whilst Prof, von Me*hely would 

 reverse the series, considering the reticulate platycephalous 

 types as the more primitive, from which the striated pyrami- 

 docephalous are derived. He now even postulates a deri- 

 vation of the Wall-Lizards from pristidactyle types with 

 transparent lower eyelids, whilst I would regard the latter as 

 representing specialization from liodactyle types with opaque 

 eyelids. Mountain forms of these lizards he regards as relics 

 of a former epoch, whilst I would look upon them, in most 

 cases at least, as modifications of the forms of the plain, in 

 the same way as Salamandra atra is surely derived from 

 S. maculosa or some form closely related to it, and not the 

 reverse. I am strongly of opinion that L. bedriagce and 

 L. sardoa are more nearly related to the forms now living 

 in Italy, Elba, Corsica, and Sardinia than to any others, and 

 bear no close genetic relationship to L. oxycephala or other 

 platycephalous forms of S.E. Europe, whilst L. monticola 

 (a variety of L. muralis which Prof. v. Mehely regards as 

 a species) is derived from the L. muralis of Spain and 

 Portugal, its similarity to L. horvathi and L. saxicola being 

 an example of convergence. 



Reserving for a future occasion a full discussion of these 

 questions, I will only observe at present that the phyletic 

 considerations which have guided the author in regarding the 

 " Schwesterformen " L. danfordi and L. anatolica as valid 

 species have not been logically applied to L. bedriagm [reticu- 

 lata of Mehely) and L. sardoa, which, from his own remarks, 

 seem to stand in exactly the same relation to each other. 

 The author, it may be noted, no longer regards L. sardoa as 

 simply identical with L. bedriago3. 



characters of the skulla rested upon the study of a large material, and 

 now I am greatly surprised to find that for 21 forms dealt with in this 

 contribution he has only examined 47 skulls altogether, and of one only 

 as many as 9 ; in 9 cases he has prepared only 1 skull, and in 3 he has 

 not seen the skull at all. Only of 4 forms has he seen more than 3 skulls. 

 This is not quite in accordance with what his previous statement led us 

 to expect, and if it is true that, as he himself says, the cranial bones vary 

 as much as any other organ, more skulls should have been studied. 



* Ann. & Mag. Nat. Hist. (7) xx. 1907, p. 39, and P. Z. S. 1908, 

 p. 1063. 



