CENTRAL AMERICAN RAINFALL. 
7 
require adjustment if we are to have amounts for which 
comparisons are significant. I have therefore adjusted the 
actual annual values to get comparable means, using both 
the periods and a well known and safe principle of meteor¬ 
ological interpolation, which runs as follows: “In the long 
run the meteorological elements of neighboring stations vary 
together and in much the same ratio.” 
To piece out my statistics I have further interpolated an¬ 
nual values when I had less than twelve but more than eight 
months’ of observations. In the case of Punta Gorda, in 
British Honduras, I have gone so far as to make an annual 
mean based on two months’ observations, an inspection of 
the place, and common report. For instance, about Lake 
Izabal, at sealevel and not far from Punta Gorda, the com¬ 
mon saying is that it rains thirteen months out of every 
tw r elve. There is no doubt that the rainfall at the head of 
the Gul'f of Honduras is very high, and the two months’ 
observation available from Punta Gorda, compared with 
Belize, enabled me to guess how high with an approxima¬ 
tion better than nothing. 
The adjusted annual values appear in the second column 
of Table III and are entered on the chart for annual rainfall, 
Plate 1. Several interesting conclusions appear at once: 
1. The rainfall is greater on the Atlantic than on the 
Pacific side as two or three to one. 
2. The greatest annual rainfall observed is at Greytown, 
Nicaragua, where it reaches the enormous total of over twenty 
feet—a figure surpassed in America only on the Mexican 
Gulf coast, in the West Indies, Guiana, and on the coast of 
Brazil. 
3. The next greatest is in Alta Verapaz, on the northern 
slopes of mountains, and on the adjoining southern part of 
British Honduras. Next to this comes the east coast of 
Panama. 
4. The region of smallest rainfall is along upper plateaus 
in most of Central America proper, but moving to the south¬ 
ern coast in the Isthmus region. The area sketched on the 
