CENTRAL AMERICAN RAINFALL. 
11 
Turning now to the lower central plateaus and the Pacific 
versant, where we have the typical invierno , we find : 
Station. 
Elevation. 
Rainfall. 
Cor into. 
Feet 
Sealevel 
200 
218 
2,156 
3,724 
4,265 
4,856 
Inches. 
90 
|}eo± 
72 
S} 64± 
54 
Rivas. 
Granada... 
San Salvador... 
San Jose. 
Tres Rios..... 
Guatemala. 
Here the interpretation is not simple, but it is safe to 
conclude that up to 4,000 feet there is very little varia¬ 
tion with height, and that the variation with topogra- 
P h y [ { Granada ( or { Tres Rios } ] is much more im P or - 
tant than that with the elevation. There is some indication 
of a maximum at 2,000 to 2,500 feet, as in Alta Verapaz. 
4. Distribution through the Year .—If we examine the dis¬ 
tribution of rainfall and of rainy days we find that there are 
four fairly distinct types, as shown by Table IV and Plates 
2 and 3. 
I. A type with typical invierno occupying the plateaus of 
moderate elevation and the Pacific versant from the north 
to southern Costa Rica and probably to the Gulf of Panama. 
It is characterized by maximum rainfall in June and in 
October, by almost no rainfall from November to April (the 
verano), and by a secondary minimum in August. The last 
is the veranillo, already referred to, and often consists of an 
almost complete cessation of rain for a week or fortnight. 
Comparing the curves of amounts of rain with that of the 
number of rainy days in the diagram, we find that what 
little rainfall comes in the verano is scattered through a rela¬ 
tively large number of days and is therefore very light, while 
that of the maxima (June and October) is scattered through 
relatively few days and is consequently heavy. This is 
