Periodical Literature. 629 



/. Errors in Diameter Measure. 



1. Errors in diameter measurements with calipers come most 

 frequently from the fact that one, or both of the arms are not 

 at right angles to the scale. This error, if the scale is pushed 

 close to the stem or log, is proportional to the angle by which 

 the arm or arms are not at right angles, and are in direct relation 

 to the true diameter, i. e. the diameters of large and small logs 

 are read off faultily in the same proportion. 



The following table gives the errors in per cent, of the true 

 diameter for different angles of error: 



angle of error o°3o' 1° 2° 3° 4° 5° 6° 



per cent of error: 40 .871.742.573.494.375.26 



If, therefore, the angle of error on one, or both caliper legs 

 is 3°, the diameter will be measured 2^% short. With this table 

 it is possible to correct caliper results obtained by a faulty caliper. 



The error is different when, as most frequently with small 

 logs, the scale cannot be pushed close to the log. Here the error 

 is uncertain, depending on the vertical distance of the scale from 

 the log. Hence, for practical work the caliper should be so con- 

 structed (short arms) as to permit close contact of scale to log. 

 If this is not done, the same faulty caliper measures the smaller 

 logs with absolutely and still more relatively greater error, than 

 the larger; and the error can be quite considerable. For in- 

 stance, for an error of angle of only 2^ and a true diameter of 

 4 inches with a distance of the scale from the log of 10 inches 

 'the error would be 8.7%, while if the scale could have been 

 pushed close to the log it would have been only 1.44%. 



2. Another frequent error arises from applying the caliper 

 so that it will not be at right angles to the plane in which the true 

 diameter lies, when the measured diameter is necessarily larger. 



The amount of error depends, of course, upon the angle at 

 which the caliper deviates from the right angle. From the fol- 

 lowing table it appears that the error is not great until about 7° 

 deviation is attained. 



angle of error, . 1° 2° 3° 4° 5° 6° 7° 8° 9° 10° 12° 16° 

 per cent, of error: .02 .06 .14 .24 .38 .55 .75 .98 1.25 1.54 2.23 4.03 



This error works in opposite direction to the first one dis- 



