12 BULLETIN 193 



ing clouds, and there is no satisfactory instrument which will record 

 accurately the light values during an entire day. As has been pointed 

 out, there are many openings in the crowns of the individual trees and 

 between the crowns of the adjacent trees, which cause an uneven shade 

 on the ground. Indeed marked variations occur within a few feet. 

 On this account great care must be taken in making exposures. In 

 the studies now under review while making exposures in woods which 

 were free from shrubs, the operator walked back and forth over a 

 distance of 50 feet. The photometer was so held that the shadow of 

 the body did not fall on the instrument at. any time. In other habitats 

 it was impossible to move the photometer through so great a distance, 

 but in no case was it held stationary while an exposure was being 

 made. Nevertheless despite precaution it is impossible to select an 

 average station, for the personal equation is bound to be an all-im- 

 portant factor in its choice. 



Absolutely cloudless days are few and far between in Michigan 

 and Vermont. To be convinced of this fact one has only to attempt to 

 make a study of light intensities. He will be struck with the fact that 

 slight clouds pass almost continuously. Light variations which one 

 would not notice ordinarily nevertheless are quite marked, even within 

 the space of a few minutes. On this account, especially in the work 

 of the past three years, simultaneous readings have been made. One 

 person was left in the open control station while others went to the 

 forest stations. If watches are compared and definite times for ex- 

 posures are agreed upon, simultaneous readings can be made, even at 

 stations which are widely separated. The lack of simultaneous read- 

 ings is a serious source of error occurring in and vitiating much of the 

 published data on light studies which are now extant. Knuchel (10) 

 recognized this limitation of accuracy in his own work but was unable 

 to rectify it because he did not have access to a sufficient number of in- 

 struments to enable him to make simultaneous readings. In the earlier 

 work, before the need of simultaneous readings was felt so keenly, the 

 author made observations at close intervals, usually five minutes apart, 

 in order to avoid as much as possible the error due to changing light in- 

 tensities in the open. 



Even the velocity of the wind has been found to introduce errors 

 in the readings taken in the forest, especially in hardwood forests. 

 The movement of the leaves is an important factor in determining the 

 amount of light passing to the interior of the crown or to the ground 

 cover. 



