Acanfhodadyhts. 61 



The presence of Algerian species of Acanthndadijlus in Somaliland, in 

 company with Paammodroiiins ahjiriis. which appears in the same list, 

 would be highly surprising from our present zoo-geographical stand- 

 point, and, although many hundreds of lizards have since Ecvoil's time 

 been received from various parts of Somaliland, and reported upon in 

 •2-2 different papers, no coutirmatiou of the occurrence of any Aratitho- 

 dacfyliis or Psammodronma has appeared. But the form named 

 A. raiUanti has since been re-discovered at Oran, and the notes 

 I)u1)lished by Lataste and by Anderson show the specimens originally 

 referred to A. vulgaris and A. savignyi to agree in all particulars with 

 forms characteristic of Algeria (A. pardalig, vars. hedriaijiv, latasfii, 

 A. scufeUatiis, var. inornafa). I have therefore not the least doubt 

 that by some mistake, for which I can suggest no exjilanation, the 

 specimens in C|uestion, together with the Psammodroiuus algirus, all of 

 Algerian origin, got mixed up with the Somaliland Collection. 



Then as to the identificatiou with ^4. savignyi, which Doumergue 

 has accepted at my suggestion. 



We have nothing to go by but the figure in the Description de 

 I'Egypte, which appears to be executed with great care, but does not 

 show the side of the head nor the full number of longitudinal rows of 

 ventral plates. The original specimen has been lost. But, with the 

 exception of the smaller interparietal, everything shown in the figure 

 agrees with the Oran lizard : the general shape, the " facies tout 

 particulier, presque autant de Latastia C(ue d' AcantJtodactylus " 

 according to Lataste's descrij^tiou of A. vaillanti, the length, of the 

 tail,* the shape of the head, the division of the frontonasal and the 

 presence of two shields between the prefrontals (which I have never 

 found in any of the very numerous specimens of A. pardalis), the 

 supraocular region with its large granular areas, the gular fold and the 

 fold of the collar, the absence of auricular denticulation, the strong 

 fringe of the toes, the strongly keeled dorsal scales, even the markings 

 on the back,t are characters which we find combined in the Oran 

 species, and in no other, certainly not in A. 2Mrdalif, to which Lataste 

 and Anderson were inclined to refer Savigny's lizard. 



One of Lataste's arguments in favour of identifying Savigny's 

 figure with A. hedriagcv {A. iMrdo-Us), was this: "A priori il faut 

 identifier est individu a uue des especes couuues d'Egypte ; car il u'est 

 pas vraisemblable que, dans un pays aussi explore, il appartienne a une 



• Twice that ot head and body; never more than 14, usually not more than 

 li times in A. pardalis. 



t "Chez les vieux individus, les bandes se seotionnent de plus en plus, et 

 certains deriennent presque pommeles." Doumergue, op. cit., p. 170. 



