392 Forestry Quarterly 



tions from it can be ascertained. For certain large groups of 

 investigations, as for instance chemical analyses, a generally de- 

 terminable percentage of error may be ascertained, and then a 

 single analysis may suffice, but in less developed and less well- 

 known fields, and where great accuracy is to be attained, repeti- 

 tion alone brings safety. 



In silvicultural experiments the results of certain treatment is 

 not only dependent on this treatment, but also on the site. On 

 another site the same treatment produces a more or less different 

 result, which may even be entirely opposite. Hence an applica- 

 tion of silvicultural experiences can only be made if the site is 

 recognized as the same or different from that on which the ex- 

 perience was secured. 



The fundamental start for a scientific silvicultural experimen- 

 tation is the question, what variations show the sites believed by 

 us as alike, when the same species in the same manner is grown 

 on it. 



The answer to this question of the variation of sites must in 

 the first place be given for the different parts of the same experi- 

 mental area, and then by parallel or control experiments. Such 

 parallel or check areas are usually lacking in forest experiments, 

 and as a rule a mere belief in equality of sites exists. Such 

 experiments cannot furnish scientifically correct data. 



It is customary to use yield tables, divided into five site classes, 

 for determining soil quality, yet they contain nothing of the detail 

 of soil, climate, topography, elevation and other influential fac- 

 tors of production. Such are good enough for cameralistic use 

 in regulating utilization, which is revised every decade, but they 

 are not fit to serve as basis for scientific, natural history conclu- 

 sions. They characterize the site by the stand that happens to be 

 on it, but how the site would fit stands of other character they 

 cannot tell. For the classification of farm lands in Saxony not 

 less than 43 field classes, each with 21 grades are used ; and for 

 meadows not less than 19. 



It is entirely erroneous to assume, as is often done, that the 

 progress of development of the stands on these as equally good 

 recognized sites in the forest really developed in the same manner. 

 From the equality in height of two stands in the same age, it does 

 not follow that the heights have been attained in the same manner 

 or that they will grow alike in future; the same with volumes. 



