106 VIEW OF THE FAUNA OF BRAZIL 



mination of which clearly explained to me the mode of their 

 introduction ; namely, that these creatures, whose footsteps I 

 had often traced in the mouth of the caverns, had undoubt- 

 edly visited them for the purpose of licking the saltpetre, of 

 which the crystals cover the surface of the floor ; and that 

 they had lost themselves in the labyrinthine passages. 



Of predatory animals there is no existing species that makes 

 caves its habitual dwelling-place. I certainly have remarked 

 an abundance of the foot-prints of the Couguar (Felts concolor), 

 not only at the entrance, but a long way in the interior of ca- 

 verns ; but I imagine it only enters them when following the 

 track of the wild hog or the paca. That it does not use the 

 caves as fixed dens, I presume from the fact that I have never 

 found the animal itself, nor the remains of its meals there, 

 while I have often had opportunities of seeing them in the 

 woods, even in the immediate vicinity of caves. 



The two remaining orders, the Edentata and Quadrumana, 

 do not enter into the question here, forasmuch as I have not 

 hitherto found the slightest trace of them in the Brazilian 

 caverns. However, the imperfect state of our knowledge re- 

 specting the first of thesd orders, together with the important 

 part they have played here in a former age, compel me to 

 take a cursory survey of the species of this class which now 

 inhabit this part of the New World. It is divided into the 

 three natural families of the armadilloes, the ant-bears, and 

 the sloths. I shall commence with the most numerous, and 

 the most difficult, the armadilloes. 



The unfortunate idea of founding the determination and 

 nomenclature of the species of this family on the different 

 number of the bands, has introduced such confusion into their 

 natural history, that most later authors have recognized the 

 necessity of beginning de novo the systematic arrangement of 

 these animals. In consequence of these attempts some of the 

 previous confusion has been removed ; but at the same time 

 new errors, in my opinion, threaten to creep in, as long as 

 naturalists will continue to determine these very variable crea- 

 tures, from individuals kept in collections. In this district of 

 Brazil there are found four species of this family ; of which 

 by far the most abundant is the Dasypus longicaudus, Pr. 

 Max. On account of its universal occurrence it is called by 

 the Brazilians, the " Tatu-verdadeiro," i.e. the true or proper 

 Tatu ; which indeed is nothing more than a translation of its 

 Indian name, Tatu-ete, which has the same signification. — 

 Marcgraaf has given a very good description of this species, 

 and under its right names, Tatu-ete, Tatu-verdadeiro ^(p.231), 

 but by a blunder of the binder, the wood engraving repre- 



