262 A YEAR OF COSTA RICAN NATURAL HISTORY 



had always found this species in deep woods and far from 

 water; here the woods were deep, but — . 



The next morning I went to these woods again but al- 

 though I spent two hours in the immediate neighborhood, I 

 saw neither monkeys nor Megaloprepus, nor indeed much of 

 any kind of animals and no dragonfly larvae in the streams. 

 So I returned to the road, which continued to ascend through 

 the forest; here and there in it were pools of rain-water, occa- 

 sionally supplemented from little running streams which 

 preceded the human-made road or have taken advantage 

 of it. I had caught a dragonfly at one of these pools and 

 was putting it away when, looking ahead, I saw an arma- 

 dillo (Dasypus novemcinctus fenestratus) crossing the road 

 some twenty feet away and disappear into the shrubbery. 

 My next visit thither was on March 24, 1910, a cloudy day 

 with fitful gleams of sunshine and I could find few dragon- 

 flies, no monkeys and no armadillos. 



Mr. H. H. Smith, when collecting in Brazil, observed 

 that certain dragonflies were only or chiefly to be found just 

 after sunset along river banks. With this in mind, I went 

 after dinner on the first day at Peralta (August 7, 1909) to 

 the Rio Reventazon about one hundred and fifty feet away. 

 Mr. Hess called a Jamaican negro to guide me through some 

 garden patches and cane-brakes, after which we crawled 

 around and over numerous boulders and so came out on a 

 flat, treeless and rocky beach. The Reventazon is moder- 

 ately rough here between rapids above and below. My 

 guide showed me the height to which the river reached in the 

 great floods of the preceding December and January. Be- 

 tween the river and the station the land rises a little and in 

 the same floods — as Mr. Hess told me at another time — the 

 Reventazon broke through its banks farther upstream and 

 came down the railroad tracks to a depth of four feet near 

 the station. Owing to landslides above and below which 



