890 JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES VOL. 11, NO. 16 



than one second if an accuracy of 2 per cent in the determination of range is 

 to be secured. To build an instrument which shall successfully attain this 

 accuracy under service conditions and in the hands of men not thoroughly 

 grounded in laboratory practice calls for the greatest refinement of optical and 

 mechanical design. 



The errors to which a range finder is subject are of two kinds. The adjust- 

 ment error is a systematic error which is constant, when measured in angular 

 units, for all settings of the instrument and may be considered as correspond- 

 ing to a bodily shifting of the entire scale of the instrument with respect to 

 the index. An adjustment is provided for correcting this error in the field. 



The dispersion error is an accidental error arising from the inability of the 

 observer to correctly determine precise coincidence in the field of the instru- 

 ment. It may be either positive or negative and its distribution for a large 

 number of observations may be expected to be in accordance with the ordi- 

 nary frequency curve for errors. This error can be studied satisfactorily only 

 by a laborious statistical method. A test has recently been completed at 

 Fort Sill by the Ordnance Department with the cooperation of the Field 

 Artillery and the Bureau of Standards which was represented by Mr. W. O. 

 Lytle. Approximately 21,000 separate readings of range were taken by a 

 group of average observers with 17 different instruments on a selected series 

 of targets. The most probable value of the dispersion error was found to be 

 of the order of 0.9 second. Much greater differences were found in the 

 dispersion errors of the different observers than in those of the different 

 instruments. This indicates that at present the selection and training of 

 observers offers a much better prospect for an immediate and spectacular 

 increase in the accuracy of range finding than does the careful selection of an 

 instrument from those types now available. For the poorer observers the 

 dispersion error was approximately twice that of the better. In other words 

 the poorer observer requires a two-meter base range finder in order to obtain 

 an accuracy equal to that of the better observer with the one-meter instrument. 



The paper was discussed by Messrs. Hawkesworth, Wright, Sosman, 

 Humphreys, and Faris. 



The second paper, on The spectral distribution of energy required to evoke the 

 gray sensation, was presented by Mr. Irwin G. Priest and was illustrated. 



The chief significance of this paper lies in the development and testing 

 of an experimental method for determining an objective physical standard 

 of "white light." The standard to be determined is the Planckian ("black 

 body") distribution of energy required to evoke the hueless sensation of 

 brilliance, commonly called "white" or "gray," under certain standard con- 

 ditions stated in detail in the paper. 



The method of producing and adjusting the spectral distribution of the 

 stimulus is this: Light from a lamp of known spectral distribution is modi- 

 fied by rotary dispersion in a system of quartz plates and nicol prisms in such 

 a way that by rotating one of the nicols the light emerging from it can be made 

 to assume the spectral distribution of a Planckian radiator at any desired 

 temperature between 4000° and 7000° K. Such a system is, in effect, a selec- 

 tive light filter of adjustable spectral transmission. 



Two methods of observation were described: (1) The method of adjust- 

 ment by trial, in which the observer himself adjusts the stimulus until he 

 calls the sensation "white." (2) The method of answers, in which the opera- 

 tor conducting the experiment adjusts the stimulus to correspond to certain 

 fixed temperatures of the hypothetical Planckian radiator; and records the 



