E. LÖNNBERG, MAMMALOGY OF ECUADOR. 77 



scaliness is obscured. Stränge to say the fourth and fifth 

 toe on the right foot are united along their whole length 

 (Pl. fig. 5) in one of the specimens. This is of course an 

 anomaly, but it may have some interest because it proves 

 how easily two toes may be united in these animals. As 

 already is known, Caenolestes has arboreal habits. Its feet 

 with their naked, warty soles and the well developed pads 

 must be useful in climbing, but the före feet undoubtedly 

 exhibit more pronounced adaptations to the arboreal life. 

 The red uc tion of the claws on the first and fifth finger to 

 nails and the displacement of the latter in direction towards 

 the carpus must be interpreted as such adaptations. Although 

 neither the pollex, nor the fifth finger are directly opposable 

 they are certainly more free in their movements than the 

 corresponding fingers in f . i. Phascologale or some other pri- 

 mitive Marsupial. They serve therefore without doubt as 

 useful grasping organs, and the fifth finger appears to do so 

 even in a higher degree than the pollex itself, to judge from 

 the fact that it is more powerful and has a better developed 

 pad at its base. 



The struoture of the hand of Caenolestes (fig. 3) is thus 

 something sui genens and it differs not only from such Mar- 

 supials as Phascologale but also from Didelphyidae. With 

 regard to the feet Caenolestes differs still more from the latter 

 which are provided with a perfectly opposable hallux. It is 

 thus already from this fact quite clear that Caenolestes cannot 

 be derived from any member of Didelphyidae or any Marsu- 

 pial organised as these. It appears more probable that it 

 has developed from ancestors with terrestrial habits, and 

 which have been adapted to such so long time that the 

 hallux already had become to some extent reduced. Then 

 Caenolestes, or some of its progenitors, secundarily aquired 

 arboreal habits and became adapted to that kind of life in 

 such a degree as it now is. The once reduced hallux could 

 not, however, develop again to a real gripping instrument, 

 it was too weak for that. Only its pad increased to some 

 degree. The före feet were, however, somewhat more trans- 

 formed, as has been described above, so that not only the 

 pollex, but also in an analogous manner the fifth digit could 

 be useful in climbing. 



