DEPARTMENT OF BOTANICAl, RESEARCH. 



6l 



The rates at Tucson, Mecca, Salt Lake City, Austin, and Laramie are clearly 

 indicative of desert conditions and denote xerophytic vegetation. The rela- 

 tive amount of water loss at Chicago, Grand Rapids, Burlington, and Orono 

 is found to be characteristic of areas in North America inhabited by conifers. 

 The greater losses at St. Louis, New York, Newark, and Raleigh point in- 

 dubitably to a forest-covering of deciduous trees. The surveys of 1908 will 

 doubtless result in the acquisition of data in which a more definitive analysis 

 of the matter may be made. 



From a local series of instruments, extending from the Desert Laboratory 

 domain at 2,663 ^^^^ ^^^ 3-095 ^^et down across the mesa-like slopes up to the 

 plantations in the Santa Catalina Mountains to the northward, it has also be- 

 come apparent that the determination of the evaporating power of the air fur- 

 nishes a very valuable criterion for the comparison of different plant habitats 

 and for the analysis of the plant population of any habitat. Thus the follow- 

 ing average weekly rate was found : 



The examination of the topographic features of the habitats tested shows 

 that the irregularity of the decrease of evaporation from lesser to greater ele- 

 vations is due to the action of hot dry winds in the direct lee of which some 

 of the stations lie. In all such cases the vegetation of the lower slopes is car- 

 ried above its average limit of distribution. 



Physiological Regulation of Transpiration. — No subject is of more impor- 

 tance in the physiology of plants of dry regions than that of water loss, and 

 Dr. Livingston has continued his studies upon relative transpiration with 

 mesophytic forms under humid conditions to obtain data for comparison with 

 results at the Desert Laboratory. One plant which has been thoroughly tested 

 shows a close connection between the diameter of the stomatal openings at 

 any time and the relative transpiration. Also, by the use of the cobalt method 

 of estimating transpiration, in connection with the atmometer previously de- 

 signed by Dr. Livingston, it is possible to determine with fair exactness the 

 total water-loss of any plant without disturbance to more than a few leaves or 

 to a comparatively small proportion of its transpiring surfaces. 



