52 Bulletin American Museum of Natural History. [Vol. XXXVI, 



presumably are responsible for the change from the open, scrubby vegetation 

 of the arid coastal area to the luxuriant forests which now almost continu- 

 ously line both shores of the river. Away from the border of the river, 

 however, at least on its eastern bank, the arid zone continues as far as 

 Puerto Nacional whence, according to Wyatt (Ibis, 1871, p. 117) "the fu-st 

 few miles" of the road to Ocana runs through "small savannas, tracts of 

 open grassy country sprinkled with a few stunted trees, or through woods." 



In the more northern part of this humid region tributary streams may 

 make their contribution to the muddy waters of the Magdalena through 

 marshy or low-lying land, but farther up the river the banks are higher and 

 the shores of entering streams are forested. 



The humid zone of the floor of the Magdalena continues with no diminu- 

 tion in the luxuriance of the vegetation as far up the river as La Dorada, 

 about 600 miles from its mouth. Between this place and Honda a marked 

 change occurs. Strongly eroded buttes with castellated outlines appear, 

 the soil is thinner and less fertile, and although the rainfall is not so low as 

 at BarranquiUa (Mr. Miller, the manager of the railway between La Dorada 

 and Ambalema, tells me that at Mariquita, a few miles east of Honda, 

 it has ranged in a few years observation from 85 to 100 inches annually) 

 the vegetation suggests that of an arid or semi-arid region. The heavy 

 forests are replaced by a more stunted growth and there are large tracts of 

 open country devoted to grazing. This condition apparently prevails to 

 the head of the Magdalena Valley. 



Honda to Bogota. — In a region which has been inhabited by white man 

 for as many years as that lying between Honda and Giradot, and between 

 these towns and the plateau of Bogota, it is often difficult to determine just 

 what changes man has wrought in the character of the country. At present, 

 however, in following either the mule trail fromHonda or the railroad from 

 Giradot, one sees but little forest growth between the Magdalena river and 

 the Savanna of Bogota. In the upper Magdalena Valley proper, the absence 

 of heavy forest, as has been remarked, is doubtless due to the character of 

 the soil, but on the mountain slopes the first-growth timber has no doubt 

 disappeared in many places before the agriculturist. Remains of this 

 forest were discovered between El Consuelo and El Alto de Sargento on the 

 first ridge -of the Andes east of Honda, where at an altitude of some 4000 

 feet, we found such characteristic species of the Tropical Zone as Formicarius 

 analis and Myrmelastes immaculatus. 



At El Vergel (alt. 5500 ft.), on the summit of the second ridge, or that 

 lying east of Guaduas, there is a small area of apparently primeval forest 

 in which oaks, some 75 feet in height, were prominent and the presence 

 here of Xanthoura yncas galeata, Brachyspiza capensis, Melanerpes flavi- 



