24 Alexander G. Rutjiven 



up to an elevation of about i,ooo feet, but never naturally in open savannah. 

 In fact, the seeds will not germinate except in damp, shaded places, such 

 as in the shade of the forest. Since these palms are so abundant over prac- 

 tically all of the savanna areas between Valencia and Camperucho, it leads 

 one to believe, together with other facts, that these savannas are not natural, 

 but also the result of deforestation and burning, partly intentional and partly 

 accidental, of a drier, more open class of woodland, which, growing over 

 a very hard, impervious subsoil, would not readily reproduce itself after 

 having once been destroyed. The argument that the palm would have been 

 destroyed by such fires is erroneous, because this palm was seen in many 

 instances burned around the base so deeply that not half of the original 

 volume of the trunk was left, yet the trees were still luxuriant and strong 

 and bearing as large seed clusters as the unburned individuals. 



After crossing the savanna-clad hills of Alto de las IMinas, we dropped 

 down abruptly into another and distinct region of forest. In some places 

 it was damp and dense, but for the most part it was rather light and open, 

 with much tangled undergrowth. The country is more undulating, the soil 

 more gravelly and less fertile, conditions which continue from Alto de las 

 Minas to the Fundacion River, except in the small valleys and on the great 

 flood-plain of the Rio Ariguani. The forests of the Ariguani are truly 

 magnificent, consisting of gigantic trees and a sparse undergrowth ; they 

 are dense and gloomy, even at mid-day. Typical humid forest types of 

 birds prevail, and one is reminded of the forests of the Sinu, lower Atrato, 

 and eastern Panama and Costa Rica. Giant creepers and lianas and the 

 peculiar parasitic fig tree abound, although epiphytes are not particularly 

 abundant. The foothills, from Alto de las Minas northward, are also clad 

 with a forest which extends unbroken upward. From the Fundacion River 

 to a point slightly north of Rio Frio the lowlands were originally clothed 

 with luxuriant humid forest, but this has been nearly all cleared away 

 between the lower edge of the foothills and the swamps which extend back 

 from the Cienega Grande for many miles, while vast banana plantations 

 and pastures have taken its place. The original forest here was in many 

 parts very similar to that of the Ariguani, but not so heavy, I imagine, and 

 with more undergrowth and wild plantains. Just beyond the Rio Frio there 

 is again a very abrupt change from fertile banana farms to desolate salt 

 plains covered with a scanty growth of acacia scrub and cacti, with a great 

 deal of a peculiar salt-bush, common at Riohaca and between Donjaca and 

 Gaira, near the sea beach. Soon after crossing the Rio Cordoba the same 

 class of dry, thorny, scrubby woodland appears which was so common 

 around Riohacha and between Valle de Upar and Fonseca. This condition 

 continues on to Santa Marta (PL I, fig.2) and around to the Rio Piedras, 

 but extending here to a considerable extent up over the lower foothills to 

 nearly 1,000 feet in many places and much more in others, especially on 

 the ridges. The valleys of the streams are well wooded, however, and the 

 acacias are less abundant than on the east and south sides of the Sierras. 

 The shores of the Cienega Grande are all low and swampy and flooded for 

 a long distance inland during the period of high water on the Magdalena 



