8 
HARRINGTON. 
map as below fifty inches of annual rainfall is based on the 
rainfall observed at Salama, in Baja Verapaz; on the state¬ 
ment quoted by Beclus from Dollfuss and Mont Serrat for 
the altos or high plains in northwestern Guatemala, the re¬ 
gion from which the streams radiate; on the common state¬ 
ment in that region that the upper part of the Motagua 
basin is the most arid part of Guatemala; and on state¬ 
ments of Squier, especially his estimate of forty eight inches 
for the upper part of Honduras. This region occupies the 
higher plateaus, but appears to extend farther down on the 
Pacific versant than on the Atlantic. I have terminated it 
near the Nicaraguan border simply because I have no evi¬ 
dence to take me further. It reappears in Costa Pica, at 
Agua Caliente, and then apparently passes over into the Gulf 
of Panama. 
To account for this distribution of rainfall we have the 
following causes of rain: The equatorial rainbelt which 
accompanies the sun in his annual journeys north and 
south, and gives rain at the station when he is in the zenith; 
the trade winds, which are here northeast and which since 
leaving the West Indies have traversed the warm Caribbean 
sea; the rains which come down on winds from the north 
after crossing the warm Gulf of Mexico; and the cyclonic 
rains, which accompany the great atmospheric disturbances 
occuringat certain seasons in the West Indies. Calling the 
first the invierno rain, the second the trade rain, the third 
the norther ram, and the fourth the cyclonic rain, we have : 
On the Pacific coast, invierno alone. 
South of the Gulf of Campeachy and north of the moun¬ 
tain ranges, 
Invierno + norther , 
or more easterly, 
Invierno + cyclonic. 
From Cape Gracias a Dios southward, 
Invierno + trade , 
and possibly for its northern part also 
+ cyclonic. 
