58 G. D. FULLER AND A. L. BAKKE 



4. The necessary number of samples is attained as soon as 

 the result becomes practically unchanged however much the 

 number of samples. is increased. How quickly this equilibrium 

 is reached will be determined by the size of the samples — -the 

 smaller the samples the larger the number required but the 

 more detailed will be the analysis. 



5. In practice the size of the sample is determined by the 

 proportional adjustment of the time consumed to the accuracy 

 of the results required. As the result of my experiments I 

 have found that 0.1 sq. m. is a suitable size, giving fairly con- 

 stant results from a moderate number of trials (25 to 50). 



6. If we investigate a formation which is essentially uniform, 

 with ill-defined limitations, the samples may be taken either at 

 random or along definitely determined lines. If we are not sure 

 that the vegetation of a formation is essentially uniform, the 

 question must be investigated, but it so happens that its investi- 

 gation is similar to that of frequency determinations because the 

 samples are, in this instance, taken in definite areas along marked 

 lines and each sample is referred to a prescribed place in a hypo- 

 thetical scheme. By this means the presence of a community 

 differing from the one under investigation is at once detected 

 by different frequency indices for the species or by the appear- 

 ance of new species with higher frequency indices. 



When a vegetation comprising several formations is under 

 investigation, the samples must be properly distributed over 

 the entire area according to the established boundaries if accu- 

 rate results are to be obtained. 



7. The sample-plots are selected and marked out most easily 

 by using a metal ring fixed upon a walking stick holding a metal 

 rod at right angles to the stick. This rod is of just such length 

 that the outer end will describe a circle of 0.1 sq. m. when the 

 walking stick is stuck into the soil. 



8. In the analysis of the sample areas species are counted 

 which have rooted living shoots within the area and also those 

 which have perennial shoots above it, even if they are not rooted 

 in the circle. 



