390 ANNUAL REPORT ©MITHSONIAN INSTITUTION, 1908. 
influence of the trade winds. Clouds are not seen passing to the 
Cordillera from the Pacific. The mists of the coast which drift in- 
land from the Pacific form at the season when the sky in the Cor- 
dillera is clear and their movements are with the land and sea breezes. 
Systematic observations of the rainfall in the Andes region have been 
carried on at only one locality, namely, Cailloma, which is situated 
north of Arequipa and just to the east of the Continental Divide. 
From the published data 
| nthe writer has constructed 
1897 1898 1899 1900 1901 “the accompanying dia- 
grams (figs. 38 and 4) 
which show the annual 
and monthly variations 
of the rainfall. 
The Cordilleras of the 
Andes. 
DESCRIPTION BY HUMBOLDT, 
1802. 
- Although Humboldt 
did not have Peru as an 
object of special study 
and did not visit the 
country excepting to see 
the coast at Pisco and 
Lima and to travel in the — 
northern highlands _be- 
tween Cajamarca and the 
Maranon,? he neverthe- 
less gave a°graphic and 
to a large extent a correct 
description of the chain 
of the Andes, availing 
himself of data furnished 
by others. He says, in sub- 
stance, that in southern 
Peru there are two branches of the Andes which include between 
them the Titicaca basin. To the north of the Titicaca basin there 
is a knot which includes Vilcanota, Carabaya, Abancay, Huando, 
and Parinacochas. After this knot of Cuzco and Parinacochas, in 
latitude 14° S., the Andes present a second bifurcation, and north- 
ward the two chains lie on the east and west of the river Jauja. 
Fie. 3.—Annual variation of rainfall at Cailloma 
during seven years. 
@Raimondi, El Peru, Volume I, page 15. 
