PACIFIC SLOPE OF ECUADOR. PERU AND CHILI. 517 



especially pages 227 to 292, plates XXX.-XXXVII.). Very few 

 specimens were available for study during the preparation of that 

 volume. 



During the Irwin Expedition of Indiana University, Dr. William 

 Ray Allen collected at Ascotan and at Calama, both in the Loa 

 River of northern Chili, and I collected along the railway between 

 Copiapo 2-]° 21' south and Puerto Montt 41° 28' south. For the 

 most part the rivers were crossed some miles from the coast but 

 at La Serena, Concepcion, Valdivia and Puerto jMontt the streams 

 were examined near the coast. I also collected in the Rio Blanco, 

 of the Aconcagua basin, a tributary of the Rio Aconcagua. A 

 series of collections was made in Valdivia and Lake Rinihue of 

 the Calle Calle basin, and between Puerto Varas on Lake Llanqui- 

 hue and Nahuel Huapi in the Argentine. ' The fact that the same 

 species were observed over and over again in different localities 

 indicates that the main features of the fauna have become known. 



While the very large series of specimens collected give us more 

 precise knowledge of the characters of the species and of the details 

 of their distribution it is a satisfaction that the general conclusions 

 reached in the volume mentioned above need very little modification 

 as the result of the new study. 



Physical Features. — Chili is a narrow strip of Pacific slope ex- 

 tending from the Rio Sama at its northern border, which empties 

 into the Pacific at 18° south, to Punta Arenas at 53° 10' south. It is 

 270 miles at its widest (on the Tropic of Capricorn) and 50 miles at 

 its narrowest (between La Serena and A^alparaiso). Between the 

 Sama and Puerto Montt, Chili is crossed by innumerable valleys ex- 

 tending from the crest of the Andes to the Pacific. In the extreme 

 north in Tacna, the valleys contain little or no water. North of 

 Antofagasta there is a considerable stream, the Loa. South of the 

 Loa for about 300 miles extends the Desert of Atacama. Between 

 Copiapo and Valparaiso there are great valleys with a dispropor- 

 tionately small amount of water (Plate IX.). 



A small stream passes Copiapo on its way to the ocean. South 

 of Copiapo to Vallenar the country is arid. At Vallenar there is 

 another stream similar to that of Copiapo and then again arid land 

 with an occasional rain in years. A few small pools of water were 



