2 ARKIV FÖR ZOOLOGI. BAND 3. X:0 12. 



forest region. Moreno and Casahindo lie in the Andes of 

 Argentina in tlie alpine region. 3500 m. above the sea-level 

 (Prov. of Jujuy), Volcan is a mountain valley on the slopes 

 of the Andes of Argentina. Chaquimayo and Vacainayo belong 

 to the primeval forest region of Peru. 



To Dr. G. A. Boulenger, who has kindly revised two 

 of the new species, I beg to express my sincere gratitude. 



Dendobrates trivittatus Spix. 



BouLENGER, Cat. iJatr. sal. 13. 144; I), mgenimns Spix, Stein- 

 DACHNER ill Yerhaiidl. zool. bot. Gesellsch. Wien 1864 jj. 257; D. 

 Jmlineli BouL, Boulenger in Proc. Zool. Soc. 1883 p. 636; D. hraccatus 

 CJope, CoPE in Proc. Am. Philos. Society Philadelpliia 1887 N:o 125 

 p. 53; D. trivittatus Spix^ D. hraccatus Cope, and D. haJmeli BouL, 

 Fk. Werner in Verliandl. zool. bot. Gesellscb. Wien 1001 p. (VM). 



One specimen, 27 mm. in length. found sitting on a dry 

 trunk, July 1904. San Fermin, Bolivia. Holmgren. 



It agrees completely with the description Steindachner 

 gives in the paper. quoted above, on some small Dendrobates 

 specimens, which he declares for certain to be yomig D. ni- 

 gerrimus Spix (= D. trivittatus Spix). It also corresponds, 

 except in a few unimportant cases, with Cope's form D. hrac- 

 catus and very much so with Boulenger's D. hahneli. 



In reality, we also find that Cope states his species to 

 be identical with another of the juvenile forms of D. trivit- 

 tatus, described by Steindachner; but, considering it to be 

 fnllgrown, he declares it to be a species apart, distinct from 

 D. trivittatus throngh the size and colour. As he does not 

 give any sufficient reason for this opinion, there is in my 

 thought no ground to maintain Cope's form as a distinct spe- 

 cies, especially as Steindachner had at his disposal a great 

 number of specimens of different ages, which, according to 

 him, evidently show that there exists a specifical relation 

 between these small white-striped and red-spotted individuals 

 and the greater, black, uniform-coloured specimens of D. ni- 

 gerrimus (== D. trivittatus). 



Just as well we could declare this specimen to be full- 

 grown, representing a new species, »D. eiicnemis» (see Stein- 



