FOREST, FISH AND GAME COMMISSIOX. 75 



3. If, at the place of measurement, there is an extraordinary thickening or other 

 irregularity, the measurement should be taken higher or lower. 



4. The dimensions should be read off while th-e arms of the calipers lie close 

 against the tree, and at this moment the caliper 'man should step close up to the 

 caliper bar. 



5. The height at which it is decided that the measurements are to be taken must 

 be held to strictly. It should be marked in some way on the clothing of the caliper 

 man, by a button, for example. According to the investigations of Grundner, a 

 German, a deviation of six inches higher or lower makes on the average a difference 

 in the basal area of 1.05 per cent. When measurements are to be taken repeatedly, 

 as, for example, in a standing experiment, a mark should be put upon the tree with 

 a scratch-awl. 



6. Ordinarily only one diameter measurement need be taken on each stem, but 

 on stems which are very eccentric two measurements may be taken crosswise, and a 

 tally kept of the mean diameter. 



According to Hesz, one tallyman and two caliper men can measure 600 trees per 

 hour (maximum 971, minimum 422) ; according to Baur 765 trees, and in one day 

 of ten hours about 7,000 trees. In the measuring done in New York by the 

 United States Bureau of Forestry, twenty-five acres have been considered a fair day's 

 work for a party of four men. In average woods a party can measure at the most 

 about five acres per hour for four or five hours, or forty acres per day of ten hours. 



For scientific work, diameter classes of whole centimeters are allowed by the 

 German Forestry Association. When the fraction amounts to .5 cm., an addition is 

 made to the preceding class. 



The United States Bureau of Forestry makes inch classes ; the New York State 

 College of Forestry, 2-inch classes. In our forests, 2-inch classes are allowable, 

 especially for trees over nine inches in diameter. This will at least keep within 

 a 2 per cent limit of error, as the author has demonstrated in Rod and Gun in 

 Canada, January number, 1902. Lack of space excludes demonstration here. 



The reckoning of the sum of the cross section areas from the diameter measure- 

 ments can be accomplished by the use of a table of areas of circles, which may be 

 found in Bulletin 20 of the United States Bureau of Forestry, Washington, D. C. 

 Such tables have also been prepared by the Germans — Kunze, Ganghofer, Pressler 

 and Eberts. Grundner has shown that reckoning the square feet to more than 

 three places of decimals, even for scientific purposes, does not obtain a degree of 

 accuracy which warrants the extra labor, and that for most practical purposes two 

 decimal places are quite suf^cient. 



