JOHNSTON: FLORA OF MARGARITA ISLAND. 173 
According to encyclopedias, the mountain peak of this end is 60 
meters higher than that of the eastern end. Thougli in the distance 
that looks very probable, it makes little difference in the character 
of the vegetation. Little of the mountain range anywhere approaches 
the height of the peak. Most of the ridges are very steep and aj> 
parently have no forest growth, whatever. Cliffs are exposed in 
numberless places. Under such conditions, although the ])eaks are 
often hidden in the clouds there can be little moisture retained and 
consequently no springs nor rivers, a condition which agrees with the 
reports of the natives. The single town of Macanao on the west 
coast is supplied with "posa" water (rainwater and seepage collected 
in clay pits) and with imported water. One large ridge with short 
spurs runs east and west along the south shore, and to the northeast 
is a series of high hills. The only plain is on the nortliwest side. 
The character of the western end as a whole then is similar to that 
of the hills and plains of the eastern end. 
The above paragraphs give some idea of the topogra[)hy of Mar- 
garita. It seems advisable also to discuss so far as possible the tem- 
perature and moisture as they affect the conditions of vegetation. 
The above description distinctly suggests that Margarita is a very 
dry island. During the time from July 3 to August 22, in 1901, 
there were only two heavy rains in El Valle each lasting much less 
than an hour, and I did not observe any light showers. This it must 
be remembered was during the rainy season on the mainland only 
twenty-seven kilometers distant. At this time conditions seemed to 
be quite normal. Orange trees and the coconut palms were bearing 
fairly well, and the valley bottoms were filled with thriving weeds. 
Observations made from Jmie 29 to September 4, in 1903, were 
very different. At this time the coconut trees were dying, the oranges 
had turned color on the tree when less than half grown, and scarcely 
a weed could be found. The island and especially El Valle was in 
a bad drought. According to the natives there had been no rain for 
two or three and some said even five years. This latter time is shown 
to be incorrect by my owti observations in 1901, but at any rate it 
is suggestive of very little rain for several years past. Records for 
1903 are as follows: — 
July 14. Heavy rain in Porlamar; very light at El Valle. 
July 17. Light show^ers at El Valle. 
August 14. Verv heaw rains on the mountain and down through 
