4 MR. G. A. BOULENGEE ON LIZAEDS 



taurica, peloponnesiaca, and Irandti, as to preclude a rigid definition. It constitutes a 

 natural association of forms more or less adapted for climbing, as a result of which the 

 digits are more compressed and more slender, and the lower surface is more even, the 

 overlapping of the ventral plates, which is very pronounced in L. agilis and L. parva, 

 having disappeared or being much reduced, the border of each transverse series of 

 plates forming a nearly unbroken straight line instead of being notched at the junction 

 of every two shields — characters Avhich increase in degree the further the advance in 

 the chains of forms which constitute the group. 



The last two sections each contain a single perfectly isolated species, from West 

 Africa and Algeria respectively, to suggest the derivation of which we cannot appeal 

 to any of the species known at present. 



Before entering on the descriptive part, I must first state by what principles I have 

 been guided in attempting to tiace the derivation of forms in this genus — and in the 

 L. muralis group in particular. 



It is not often, when having to deal with the phylogeny of existing species, that one 

 can point to any, actually living at the present time, as forming part of the probable 

 ancestral stock ; and yet, in this case, I feel pretty confident that L. agilis and its 

 close ally L. parva have preserved the primitive characters out of which the series 

 represented in Sections I., III., and IV. of the above classification have been evolved. 

 L, agilis is a widely distributed species, now ranging over the greater part of Euro])e 

 and a considerable part of Northern and Temperate Asia ; it is highly variable both 

 in its lepidosis and in its markings, and, even without imagining a gi'eater amplitude 

 of variation than is known in the existing individuals, we find in it a combination of 

 characters which realise the ideal archaic type leading, through more or less broken 

 chains of forms, still in existence, to the most extreme modifications to be found 

 in the three groups mentioned. 



So far as the very scanty paleontological material allows us to judge, the genera 

 Lacerta and Nucras are the only representatives of the family Lacerlida; known to 

 occur as far back as the Oligocene and the Ivliocene *, and these two, which are inti- 

 mately connected and barely separable, must be looked upon as the original ancestral 

 types, out of which the allied genera Latastia, Acanthodactylus, Psammod ramus, 

 Cabrita, Ophiops, Eremins, etc., have been derived by a series of modifications which 

 may be formulated as follows : — 



1. Reduction and disappearance of the teeth on the palate. 



2. Flattening and weaker ossification of the skull (reduction of the postfronto- 

 squamosal arch), together with elongation and acumination of the rostrum, accom- 

 panied by approximation of the nares to each other (reduction in width of the ascending 



* Filhol, Ann. Sei. Geol. viii. 1877, p. 269; De Stefano, Atti Soc. Hal. Sc. Nat. slii. 1903, p. 412; 

 Klebs, Schiift. Thj-s.-oek. Ges. Kouigsberg, li. 1910, p. 2-11. 



