ELIMINATION 



183 



Belfort Observatory, Ealtimore, Md. The Draper forms are made at less 

 cost by the Draper Manufacturing Company of New York City. All of 

 these instruments are prone to work out of true and should be standardized 

 often by comparison with standard thermometer, wet- and dry-bulb thermom- 

 eters, and mercurial barometer respectively. Nor do they record with minute 

 accuracy, though they are correct enough for most purposes. 



Photometers. No form of light recorder suitable for physiological uses 

 yet exists, for the meteorological burning-glass type and black -bulb-ther- 

 mometer types record only variations between sunlight and cloudiness. 

 Clements has described a photometer, utilizing the darkening of photo- 

 graphic paper ("Plant Physiology and Ecology," 73), but it is laborious of 

 operation and does not yield a graph. Accordingly in my own laboratory 

 I have made use of an empirical system as follows. The percentage of 



Fig. 47. — The Richard barograph; Xk- 



From the Price-list of Richard Freres, Paris. 



light of full sun at noon in this latitude, in comparison with the full light of 

 the sun overhead taken as 1, is determined for each day of the year; then 

 a curve for each day is constructed by drawing a parabola through the hours 

 of sunrise, noon, and sunset, and this is the curve of full sunlight. For cloudy 

 days this curve is lowered after the following system, viz., for fleecy clouds 

 or haze, to 90% of the full amount; for heavier clouds alternating with blue 

 sky, 85%; for light continuous clouds, 80%; overcast dark clouds, 60%; 

 heavy rain, 50%; heavy snow, 40%. The light record is thus drawn as a 

 graph on the same horizontal scale as the records of thermograph, hygro- 

 graph, and transpirograph. A fuller description of the details, together 

 with an illustration of the use of these graphs, will be found in Miss 

 Clapp's paper cited under Materials earlier (page 172). The subject of 

 light measurement, from another point of view, is treated by Wiesner in his 

 new work, "Der Lichtgenv.ss der Pflanzen,'' Leipzig, 1908. 



While the foregoing method, viz., the comparison of records 

 of contemporaneously registering instruments, solves the pres- 



