468 PROCEEDINGS OF THE NATIONAL MUSEUM vol. i04 



In contrast, the autumnal southward journey of North Temperate 

 Zone migrants is begun in the face of increasingly unfavorable condi- 

 tions: lessening hours of daylight, decreasing food supply, and lowering 

 temperatures. 



It is interesting to note the very slight morphological differences, 

 as opposed to great behavioristic and possibly great physiological 

 differences, between some migratory and nomnigratory races 

 of the same species, for example: Dendroica petechia aestiva and 

 rujopileata; Vireo virescens virescens, vividior, and chivi; Capella 

 gallinago delicata and paraguayae; Muscivora tyrannus tyrannus and 

 monachus. Where a species breaks up into highly migratory and 

 relatively sedentary races, an experimental, comparative study of the 

 physiological differences should prove of great interest, particularly 

 as the breeding cycle has apparently remained, or has evolved to be, 

 about the same. 



Although a clear-cut external stimulus capable of initiating migra- 

 tion appears to be lacking, it seems that the annual cycle of wet and 

 dry seasons is an important factor affecting breeding. From Decem- 

 ber through April the woods become increasingly bare and parched 

 until the advent of the rains of May (often the latter half) and June 

 causes the woods to leaf out again. At first glance it might seem 

 that the breeding species curve closely approximates the rainfall 

 curve, as does the upward curve of trees in leaf (fig. 106). However, 

 the upward curve of breeding species precedes those of rainfall and 

 vegetation. Both November and December rainfall is higher than 

 that of February, March, and April, and much surface water is still 

 present, whereas breeding activity is much greater in March and 

 April, although the climate and soil have become extremely dry. 

 In fact, the April peak represents many forms nesting in arid, still 

 barren woods which will not be well in leaf before June. 



During March and April the junior author made repeated trips by 

 air from Cantaura (Anaco) to Caracas, both direct by way of VaUe 

 de La Pascua, and coastwise by way of Barcelona. With minor ex- 

 ceptions, he found the area covered with sparse, seasonally deciduous 

 forest, the aridity and generally barren appearance of which was very 

 marked during March, April, and at least the first half of May. In 

 the region between Cantaura and Barcelona, the dominant tree is the 

 "pui" {Tahebuia serratifolia), which, with the less common "araguaney" 

 {Tahehuia chrysantha), makes up an estimated 75 percent of this 

 woods. From the air it appears that it continues to be the dominant 

 tree between Cantaura, Valle de la Pascua, and the foothills of the 

 coastal mountain range before Caracas. On the whole, these Tahehuia 

 do not come into leaf before June. Throughout the area observed on 

 these flights at least 90 percent of the woods was leafless except at 

 the rivers (generally without surface water) where a narrow line of 

 trees in leaf paralleled the water courses. More rarely, there was a 



