APPLICATION OF PHYTOSOCIOLOGICAL TECHNIQUES 299 



ceae) and Inga (Leguminosae) have 10 species each. Two genera have five 

 species, five genera have four species, seven genera have three species, 28 

 genera have two species, and 102 genera have one species each. If it is assumed 

 that the specimens not identified beyond the family are each monotypic in the 

 flora, the latter number is increased to 126 entities. 



Concluding statement. We beheve that this study of the Mucambo plot 

 shows that certain phytosociological concepts and methods can be applied 

 to the complex vegetation of the humid tropics much as they have commonly 

 been employed in the study of temperate vegetation, with considerable im- 

 provement in understanding and description of such tj^es as rain forest. 

 Our work was greatly facilitated, in fact made possible in the short time at 

 our disposal, by the help of local mateiros, for the richness of the flora presents 

 problems of recognition and identification not common in temperate work. 

 Furthermore, there are many parts of Brazil where such help could not have 

 been received and where the flora is less well known. 



PART III. LIFE-FORM AND LEAF-SIZE CLASSES OF THE 

 BRAZILIAN RAIN FOREST « 



In a long series of studies on statistical plant geography by Raunkiaer 

 (brought together in English translation, 1934), considerable attention was 

 given to life forms of plants in the analysis of floras and communities and 

 as an approach to phytoclimate. Less attention was paid to leaf-size diag- 

 nosis, although a series of easily manipulated classes was developed. Raun- 

 kiaer's system of life-form classification met with considerable success as 

 measured by its world-wide use since it was first introduced in 1904. The 

 leaf-size classification, although equally amenable to statistical treatment, 

 has been little used. A review of the world literature on life forms and phyto- 

 climate was published by Cain (1950). 



During the latter half of 1955 and early in 1956 the authors had an op- 

 portunity to travel considerably in Brazil in connection with the project 

 mentioned in the introduction. Among the concepts and methods of plant 

 sociology that we were able to test on Brazilian vegetation were Raunkiaer's 

 systems of life-form and leaf-size classification. This paper gives the data 

 we have on the rain forest (except for some studies of the temperate rain 

 forest with an overstory of Araucaria, the Parana pine). It cannot be said 

 that our data are adequate for characterization of the life forms and leaf sizes 

 of the Brazilian rain forest as a whole, but they come from widely scattered 

 stations and are nearly complete for the flora of the sample plots studied. 

 It seems to us that these fragmentary results are interesting and sufficiently 



''' Stanlev A. Cain and Gustavus M. de Oliveira Castro. 



