590 JOURXAL OF FORESTRY 



•To my mind however, the third system mentioned above is the best 

 compromise for an ordinary extensive cruise. That is, to deal with 

 the stands in categories or groups. This is really the case where 

 cetain fixed types are adopted and the stands assigned to the type 

 that fits them most nearly, or else where fixed types are not used but 

 each stand is described individually. I prefer to use this latter system 

 since the grouping comes in only as far as the volume computation 

 and other calculations are concerned, but each stand is characterized 

 on its own merit. 



One assumes that there are two, three or even four variations or 

 classes occurring for any well-defined type of forest. The stands are 

 each assigned to any one of these classes by the estimator in the field, 

 according to whether the trees are relatively tall, medium, short, or 

 scrubby. This same idea is borne in mind in the observations on 

 sample trees, these being ascribed to their proper class. 



To make this matter more clear, let us consider a specific case. 

 Supposing a section was to be cruised where a certain kind of mixed 

 hardwood or coniferous forest was abundant. This forest type might 

 be found existing in very fertile locations where the trees grew rela- 

 tively tall. Again it might be found on side hills where growing con- 

 ditions were fair and the trees were of average height. Lastly, the 

 mixed tyi^e might exist on mountain tops where the exposure and 

 other climatic conditions caused the trees to be somewhat short or 

 stunted. (Jne could here, conveniently define this mixed wood forest 

 in classes 1, 2, and ."> ; the finest stands being put in class 1, the average 

 stands in class 2, and the exposed stands in class 3. These classes 

 might be called "volume classes." 



Use of Height in Classifying for Voluiuc 



Height, which is a certain criterion of volume, can conveniently be 

 used in ascribing each stand to its class. It is, in fact, quite common 

 to look over a region before a large estimate is undertaken and to 

 draw up a tentative height curve for each such class. The final results 

 are worked up to give their own modified height curve for each class, 

 as well as their form class curves, etc. Then, again applying these 

 results to Jonson's Form Quotient tables, we work out the volumes, 

 the number of logs, etc., as before. Of course there are certain per- 

 sonal errors creeping into this procedure but their influence is limited 

 to the placing of a stand in one class or another; the difiference in 

 volume between any two classes need not be very great. Barring 



