282 CAIN, DE OLIVEIRA CASTRO, PIRES, AND DA SILVA 



hectares, divided into 20 strips each 10 X 100 m. When we first visited the 

 Institute in September, 1955, we were taken to the plot by Dr. G. A. Black, 

 as Dr. Pires was at that time working in the United States. It was immedi- 

 ately apparent that the plot offered a fine opportunity for concentrated study 

 of a small tract of equatorial rain forest, as each tree of the 20 strips had 

 been numbered by a metal tag and careful identifications by personnel of 

 the Institute were nearly completed. Also, most of the tree species were rep- 

 resented by collections from Mucambo in the Institute herbarium. Contact 

 was made with Dr. Pires, and we returned to Belem to study the plot in 

 January, 1956. 



The original lists of Dr. Pires consisted of trees on the plot that were 

 about 1 dm. d.b.h. or more, divided into two size classes, those under and 

 those over 40 cm. d.b.h. As the trees were numbered consecutively in each 

 strip, it was possible to go onto the plot and determine from the tags which 

 stems were in each 10 X 10 m. quadrat. It was thus possible to use the data 

 for density and frequency studies. We also went back over the plot and 

 determined the exact diameter of each stem by use of a steel tape, which 

 allowed us to compute basal area for each species and for the plot as a whole. 

 Determination of the maximum height of each species allowed assignment to 

 Raunkiaerian life-form classes. Each of the 200 quadrats was searched for 

 non-arboreal species, and collections were made for determination and leaf- 

 size classification as well as for life-form classification. These various aspects 

 of the composition and structure of the forest on the sample plot will be dis- 

 cussed separately, but first the total known flora is presented in table 4. 

 Primary arrangement of the species in this table is by Raunkiaerian life-form 

 classes, and within them, by Raunkiaerian leaf-size classes. To save repeti- 

 tion this table also shows whether the leaf is simple or compound and gives 

 the blade or leaflet area in square centimeters. Common names are not used, 

 as in most cases we can provide a Latin name. Because there is no manual 

 of the flora of the region, we have added name authorities and family names. 

 Some material that has not been named is nevertheless used for life-form and 

 leaf-size study since the more complete the flora the more valuable the 

 spectra. 



Density of trees. Table 5 shows the density of stems on sub-plots of 

 different sizes formed by combinations of contiguous 10 X 10 m. quadrats. 

 This table shows the number of sub-plots in each size group that contains 

 a given number of stems 1 dm. or larger in diameter. For example, in the 

 10 X 10 m. series, 7 quadrats had only 2 stems, 14 plots had 3 stems, 32 

 plots had 4 stems, etc. The total sample of 2 hectares contains 1,188 trees 

 whose stems are approximately 1 dm. d.b.h. or over, which is an average of 

 16.8 sq. m. per stem. In order to have some measure of the reliability of the 

 density data, table 6 was developed. Using the smallest subdivisions of the 

 20,000 sq. m. plot, the 200 10 X 10 m. quadrats, it was found that the 



