324 Forestry Quarterly. 



tions, the latter by groups or in circle method, noting at same time 

 the exploitable trees, ascertaining actual increment and also at- 

 tempting comparison with normal groups, and a French method 

 which he considers the most serviceable and describes in detail. 



This method, mcthode dii controle, was invented by Gurnard 

 and worked out by Biolley, a Swiss forester, at Couvet, Canton 

 Neuenburg, on about 6,000 acres of communal and in part on 

 15,000 acres of private forest, mostly fir and spruce. In the com- 

 munal forest of Couvet, the introduction of a real cultural selec- 

 tion forest dates only from 1890, hence it is not yet all in proper 

 condition. 



Before describing the method the author brings a lengthy pane- 

 gyric of the properly conducted selection forest as exhibited or 

 aimed at in Couvet and explains the divergence of opinions by 

 pointing out that such a great variety of conditions may exist in 

 the selection forest. He brings a table comparing production in 

 compartment timber forest and in the Couvet stands, which would 

 make it appear that the increment is larger in the selection forest 

 and the use per cent, with 3.5 against 3.3 not less. 



According to Gurnaud's method the forest is to be divided into 

 compartments, a thing which for selection forest is supposed not 

 to be necessary ; only in this way can good statistical data for the 

 control be secured. The compartments are made with reference 

 to site and stand difference of not more than 25 acres extent, and 

 the areas carefully computed. 



Dimension and not age is used for classification, and rotation is 

 not a term applicable, since age and dimension do not go parallel 

 here. An inventory is, however, necessary, and a repeated stock 

 taking at that, besides an accurate accounting of felled material, 

 and the stock taking is done over the whole forest in the same year. 

 All trees of over 7 inches diameter are calipered in 2 inch classes, 

 the place of measurement being marked so that they may be 

 measured in the same place again. This is the material principal, 

 the smaller growth or materiel accesoir being left out of consider- 

 ation or being added with 10 per cent, of the principal volume. 

 The diameter classes are grouped into three size classes — small, 

 middle, and stout. 



The volumes are computed from special volume tables which 

 take into account the great variation in taper and give the ap- 

 proximate contents. This designated sv is different from the 



