32 Lacertidse. 



become reduced or disappear entirely through breaking up into scales. 

 The upper temporal shields, primarily two in number, are deeper in 

 L. agilis, and also in L. viridis, than in any other species, and are 

 situated partly on the upper surface of the head (where the anterior 

 forms a suture with the fourth supraocular), partly on the side, thus 

 combining the two' extreme positions met with in L. muralis and 

 allies, which Mehely explains as due to the shields not being homo- 

 logous. I regard them as certainly homologous : if lateral in position, 

 they have been reduced in width and pushed aside by the greater 

 lateral extension of the parietals ; if dorsal, the lower portion has dis- 

 appeared through disintegration. Mehely would agree, I should think, 

 that either case must be a reduction from the condition in L. agilis. 



I conceive five to be the original number of shields on the upper lip 

 to below the eye, the fifth being the subocular. This subocular 

 becomes more and more differentiated from the labials proper by 

 narrowing inferiorly, and may ultimately be excluded from the labial 

 border, as in some Eremias and Acanthodaetylus. In L. agilis this 

 shield is very variable in shape and is usually preceded by four upper 

 labials. Five or six anterior upper labials become normal in several 

 forms of the L. galloti and muralis groups, and the number is often 

 reduced to three in the L. vivipara group. As a general rule, the 

 number of labials increases with the length of the snout. 



6. The lower eyelid is opaque, usually with more or less enlarged 

 scales in the middle, in all species of Lacerta but one (L. perspkiUata). 

 In L. parva, danfordii and dugesii these large scales have a tendency 

 to become translucid. But in L. perspicillata a perfectly transparent 

 disc, formed of a single large scale, occupies the centre of the lid. We 

 know of no connecting-links in the genus Lacerta leading to this 

 remarkable feature, but we can realise the process of formation of the 

 disc by examples drawn from the genera Latastia and Eremias, in 

 which we find a varying number (two or more) of central scales 

 becoming enlarged and transparent, and by their fusion realizing the 

 condition in L. perspicillata. In Gabrita the transparent disc is very 

 large, occupying nearly the whole of the lower eyelid, and, a step further, 

 in OpUops, the lower eyelid has lost its mobility and fused with the 

 upper, such Lizards having been regarded as deprived of eyelids. 



6. A denticulation formed by projecting, more or less pointed 

 scales in front of the ear-opening is known in a single species of 

 Lacerta, L. atlantica, but I find a slight indication of it in some 

 specimens of L. muralis, var. campestris, and in L. jayakari. The 

 character is further developed and reaches its highest degree in species 

 of Acanthodaetylus. 



