258 



BOTANICAL GAZETTE [apRII- 



ith 



within the prescribed limits; then it was weighed again, the results 

 being reduced to an hour basis. The Hght was kept bright and 

 diffuse by manipulation of a white curtain. The value of the light 

 was considered about 90 per cent, of that received by the plant under 

 ordinary conditions. At times when the sun did not cast a shadow 

 both plants were exposed to the same intensity. To keep the plant 

 under conditions of aeration as nearly normal as possible, the case 

 was opened every morning (when the plant was watered) and the 



air blown out by a large electric fan. 



To allow the transpiration of these plants to be compared ^^ 

 one another, and especially to permit the transpiration of different 

 sizes and kinds to be compared, it was necessary to reduce their trans- 

 piration values to a common standard. For this I adopted the unit 

 system of grams per hour per square meter of leaf surface, which for 

 convenience may be designated the M^HG standard. Leaf area 

 was taken for one side only and petioles were ignored unless the blade 

 extended along the edge, when an approximate value of half the petiole 

 was added. This system has great advantages in convenience over one 

 in which both leaf surfaces are estimated, or in which the stems and 

 petioles are taken into account, and since in general in any given km 

 of plant the amount of petiole and stem surface must be approximately 

 proportional to the leaf surface, no error is involved in a comparative 

 study. In certain studies, involving structural relations, etc., e 

 stem and petiole areas would have to, be considered. The suriac 

 area was computed by tracing the leaves on homogeneous paper, an 

 weighing them after they were cut out ; this value was then reduce 

 to a square meter of surface, by the familiar method, viz., from tne 

 determination of the weight of a known unit area of the paper. Photo- 

 graphs of the plants kept under ordinary greenhouse conditions were 

 made, all upon the same scale and against the same backgroun^ 

 (ruled in square centimeters). These are reproduced with their 

 respective curves below, with the exception of three, of which the 

 plates were accidentally destroyed. 



The results of the study I have endeavored to express in the wa} s 

 which will render them most useful. First in importance are the 

 exact figures, which are contained in the following table and are 

 reduced to average per hour. In the nlant under sreenhouse condi- 



