EXPERIMENTS WITH WATER CULTURES. I 5 



are provided with flat cork stoppers one-half inch in thickness. 

 Each bottle is covered with opaque black paper or is painted ex- 

 ternally with asphaltum varnish or other opaque covering" to pre- 

 vent the access of lig-ht to the roots. The stopper is prepared by 

 cutting ten uniformly spaced vertical wedges from its lateral 

 surface, these being about one-eighth inch broad and of some- 

 what greater depth, and extending from top to bottom of the 

 stopper. Each cork wedge, after being cut out, is truncated at 

 its inner angle by the removal of enough cork to allow it to be re- 

 placed in position after a seedling has been placed in the groove 

 from which the wedge was cut. The stem of a seedling is placed 

 in each of the ten grooves, the seeds being just beneath the lower 

 surface of the stopper, and the wedges are pressed into position. 

 They should wedge the stems into place just firmly enough to 

 hold them when the stopper is inserted in the bottle. After all 

 are in position a rubber band is placed around the stopper about 

 one-eighth inch from its upper surface, to hold the wedges in 

 place. The stopper is then pressed firml\- into the neck of the 

 bottle, the latter having been filled to the shoulder with any 

 desired solution. 



If the culture is properly set up all joints are so tight that 

 practically no opportunity is ofifered for the direct evaporation of 

 water from the bottle. Either the solution should be renewed 

 or the bottle should be replenished with distilled water every three 

 or four days, according to the character of the experiment, to 

 make up for loss of water by the transpiration of the plants. 



Such cultures of wheat can be continued for three or four 

 weeks or even longer without difficulty. Since the water loss 

 therefrom is only through the plants, the entire cultures may be 

 weisfhed at each change and the sum of the losses recorded mav 

 be taken as the total loss by transpiration of ten plants for the 

 period of the experiment. This total water-loss, being propor- 

 tional to the area, and hence to the size of the leaves, a convenient 

 method is thus offered for determining the relative growths of 

 several cultures. f At the end of the experiment the green and 



t Livingston. B. E., Relation of transpiration to growth in wheat. Bot. 

 Gaz. 40: 178-195. 1905. 



