October 23, 1914] 



SCIENCE 



605 



but little evolution Las talsen place. The 

 difference from the fauna to the north is the 

 difference in species; the genera are in most 

 cases of wide distribution. 



Few species occur in both faunas, and these 

 are such widely distributed barrier-surmount- 

 ing fishes as Bhamdia, Hoplias and Lebiasina. 

 For them the Esmeraldas ridge has not been 

 a barrier. Sternopygus macrurus may have 

 come from the north; it is the only Gymnotid 

 found south of Esmeraldas. Sub-andine forms 

 also occur indiscriminately in both regions. 

 The four distinctive Pacific-slope genera are 

 confined to this region. It is a very old fauna 

 and an extremely meager one. Astyanax 

 fesiw, Astyanax or Bryconamericus Irevirostris 

 and Bryconamericus peruanus, all lowland 

 forms, confined to this area are more inti- 

 mately related among themselves than they are 

 to the nearest geographical members of their 

 respective genera. Either long isolation in a 

 region offering few environmental units has 

 caused them to converge or they have only 

 recently arisen. 



This fauna has evidently come from the 

 east coincident with the first stages in the rise 

 of the Andes. This possibility was suggested 

 by Dr. Eigenmann^ in 1909, chiefly to account 

 for the presence of Cetopsis occidentalis which 

 has near relatives in the Upper Amazon. A 

 ridge or spur of the Andes forms the water- 

 shed between the systems of the Esmeraldas 

 and the Guayas. On it is situated Santo 

 Domingo de los Colorados at an elevation of 

 1,500 feet, but its height nearer the sea is 

 further decreased. This seems to separate the 

 two faunas. It is, according to Wolf,^ of early 

 tertiary formation and probably arose but 

 slightly later than the beginning of the up- 

 heaval of the great western chain. The trough 

 in which now flow the Guayas and the Daule 

 is of subsequent alluvial formation. Deposits 

 showing tertiary depression occur at Loja and 

 in the headwaters of the Catamayo. Through 

 this route possibly have entered the fishes. 

 This point now raised more than seven thou- 



3 Eeports of the Princeton University Expedi- 

 tions to Patagonia, Vol. III., 1909, p. 361. 



•* ' ' Geografia y Geologia del Ecuador, ' ' Leipzig, 

 1892. 



sand feet above the sea is, however, the only 

 break in the majestic wall of the Andes. 



" The distribution of the GlanduUcaudince 

 shows a strange relationship between the 

 faunas of Transandean Colombia and south- 

 eastern Brazil. This relationship is confirmed 

 by the distribution of Sdlminus and other 

 fishes. The similarity is not confined to posi- 

 tive resemblances, but a number of types ab- 

 sent from northwestern Colombia are also 

 absent from southeastern Brazil and Uruguay. 



" The few genera of small fishes which suc- 

 ceeded in crossing or circumventing the 

 Eastern Cordilleras of Venezuela and Colom- 

 bia have undergone a remarkable radiation in 

 Colombia. These Cordilleras very probably 

 become a barrier before the evolution of the 

 electric-eel, the Serrasalmoninse and many 

 others of the common Amazonian sub-families 

 which are absent from Colombia. This makes 

 very desirable a knowledge of the fauna of 

 the region about Lake Maracaibo, where the 

 university later plans to send an expedition." 



The factors of vertical distribution are to 

 be considered in a region so mountainous as 

 that contiguous to the Andes. Of interest 

 here is the recurrence of the same species at 

 similar altitudes, in widely separated localities, 

 a fact pointed out by Sir Edward Whymper 

 among insects. The same species of Hemi- 

 hrycon was taken at Sandona, in the Pacific 

 slope of the Western Cordilleras; at Ibague in 

 the Central Cordilleras and at Guadual, in the 

 Atlantic slope of the Cordilleras east of Bo- 

 gota, in each case at an elevation of some four 

 thousand feet. Bryconamericus caucanus 

 occurs in the Upper Cauca and at a similar 

 altitude in the Upper Patia. Arges cyclopium 

 occurs in all the high inter-andean valleys of 

 Ecuador and is represented at Toche in the 

 Central Cordilleras nearly four hundred miles 

 to the north by a very similar if not identical 

 form. This species, known as "Humboldt's 

 fireproof fish" is the only true andean fish; 

 in the great Andes of Ecuador it ranges as 

 high as eleven thousand feet. 



At elevations not greater than five thousand 

 feet, a subandine fauna is encountered con- 

 taining fishes such as Pygidium, Arges, Hemi- 

 hrycon, Bhoadsia and Piabucina. The number 



