.■MRTHOD OF SA^FPLE PLOTS. 



67 



In the first case it is necessary to ascertain the areas, and 

 in the latter the number of trees, in both sample plot and wood. 

 As, however, the comiting of all trees gives hardly less trouble 

 than measuring them, the second method yields only a small 

 saving of labour ; it would be followed only, when the area of 

 the wood is not known, or cannot readily be ascertained. 



2. Selection of Sample Plots. 



The proportion given above will hold good only, if the sample 

 plot represents a fair average of the whole wood, so that it can 

 be considered as a model of it ; in other words, if a measure- 



Fio-. 2: 



ment of the trees on it yields an average basal area of stems 

 per unit of area, an average height and the same form factors. 

 Hence, the sample plots must be selected accordingly. 

 Here several cases must be distinguished : — 

 (a.) The quality of the wood is the same throughout the 

 area. In this case, the sample plot may be selected 

 anywhere, as long as the density of stocking represents 

 an average. In very large woods, it may become 

 desirable to take several sample plots and calculate 

 the mean. 

 {h.) Several qualities occur, which are clearly separated 

 according to area. Here each quality is treated 

 separately and one or more sample plot taken in each 

 part (Fig. 26). 



f2 



