312 BOTANICAL GAZETTE [april 



Experiment 3 consisted of only two culms; therefore the dis- 

 agreement found in the other experiments between the culms in 

 the same pot is not noticed here. 



Discussion and conclusions 



Atmospheric pressure and growth. — Much work, with con- 

 flicting results, has been done upon the effect of varying degrees 

 of atmospheric pressure upon growth. There is a rather general 

 agreement that a certain decrease in pressure accelerates growth, 

 but a difference of opinion as to whether this is due directly to 

 diminished pressure or to a decrease in the partial pressure of the 

 oxygen (8, 11, 15), also whether increased growth in water is due 

 to decreased oxygen pressure or to some other factor. 



Low pressure (10-20 mm. in experiment 1) seemed to have a 

 general depressing effect upon growth in Scirpus, as will be seen 

 in fig. 3 and table XII. The curve shows that this effect was not 

 constant, however, and that growth did not follow the variations 

 in pressure. In general there is greater growth in both control and 

 experiment during the day. The graphs for experiments 2 and 3 

 show this last point somewhat more clearly, and also show the 

 closer agreement of the two curves as the pressure in the experiment 

 was increased, but they are not sufficiently striking to merit inclu- 

 sion in this report. The power to grow fairly well with such a 

 small supply of oxygen evidently enables Scirpus to grow in very 

 poorly aerated situations. 



Diaphragm distance. — If low atmospheric pressure had any 

 effect upon distance between diaphragms, there would be a pro- 

 gressive increase, or decrease, in the experimental culms as com- 

 pared with the control, because there is a normal tendency to 

 increase the distance from tip to base. Tables VII and IX show 

 that, while the experimental culms had a greater diaphragm 

 distance throughout than the control, this did not increase pro- 

 gressively, but varied in the different regions. In experiment 3, 

 however, the results were different. Before the experiment the 

 control had the greater diaphragm distance (1 .61-1 .22 mm.)- 

 This persisted, but in a diminishing amount (table XI) , which really 

 means a progressive increase in the experimental culm. It is 



