8 ARKIV FÖR ZOOLOC4I. BAND 0. N:0 9. 



and lower labials, and mental are blackish. The loreal region 

 is much paler t han the temporal. The upper parts of the 

 limbs are darker brown than the body, dotted with yellowish 

 markings. The tail is brown, speckled with yellow, and in the 

 distal part the markings are gradually arranged to form alter- 

 nating (reddish) brown and yellowish crossbands. 



The presence of the reddish tint reminds about the colour 

 of T. rufescens as severai other characteristics do as well, and 

 T. duséni appears thus in some respects to form a connecting 

 link between T. teguixin and T. rufesceiis, although it is very 

 distinct from both. A very important distinguishing charac- 

 teristic is the great difference in the proportions of the fingers 

 and toes in the new species and the two others. My friend 

 Dr. BouLENGER lias namely kindly informed me that »the 

 digits of T. rufescens do not differ in shape or proportions 

 from those of T. teguixi7i», and it has already been shown 

 how widely T. duséni differs from T. teguixin in this respect. 



The structural differences of this new species indicate, 

 without doubt, different biological habits. The long tingers 

 and toes of T. teguixin and its sharply pointed claws are to be 

 regarded as an adaptation to a life in forests and amongtreesf s 

 these qualities must facilitate the climbing power of the lizaid. 

 On the other hand the short fingers and toes of T. duséni indi- 

 cate it to be a ground lizard, the digging power of whieh is 

 proved by its thick fingers and stout claws. 



The broadness of the jaws, and their strong mu&culature 

 as well as the strength of the neck appear to indicate a different 

 diet of T. duséni. The contents of the stomach of this single 

 specimen does not give any explanations, how^ever, of what kind 

 of food w^hich requires such strength of jaws and neck as they 

 consist of grasshoppers and larv?e. Professor Chr. Auri- 

 viLLius has kindly examined the entomological specimens 

 found in the ventricle and commmiicated the following: »The 

 contents of the ventricle of a Tuj)inamhis-\i7.aYå from Paranå: 

 4 grasshoppers, 1 katydid, 1 large larva of a Saturniid, probabl}^ 

 of the genus Hyperchiria, provided with long, branched spines, 

 1 naked larva of a Noctuid, 14 larvae belonging to the family 

 Megalopygidse 8 of which are only little hairy and light- 

 coloured, 3 strongly hairy and light-coloured, 2 densely beset 

 with black hairs, and with speckled tufts of hair. All these 

 are provided with stinging setcT so that it is dangerous lo 



