﻿92 , BOTANICAL GAZETTE [February 



interval before the next change of solutions. The volume of 

 water in these aspirators was so great that the surface sank very 

 slowly ; and with a siphon over two meters long the loss of the 

 effective pull was so slow that it could have had but little influ- 

 ence on the result if neglected; but even this I tried roughly to 

 correct by' lengthening the siphon as the day passed. 



The siphon was a glass tube with the largest bore, almost 

 ^mni^ with which the stream would be certain to carry down any 

 bubble forming in it. The siphon could carry water from the 

 aspirator so rapidly that the limited rate at which air enters the 

 latter has no great influence on the pressure in it, and almost the 

 whole pull of the siphon is exerted against each of the glass 

 stop cocks. Any attempt to regulate the flow in the usual way, 

 by restricting the siphon, reduces the suction in the aspirator; 

 then the more rapid movement of air through one culture, the 



slower it is through the other. 



ly 



ble to balance the resistance in the two parts of the experiment 9 



so perfectly that they would keep the balance through the whole 

 time of an experiment, except in the way described, by having 

 the same high and steady suction act constantly on both. 



The baryta water was put into the absorption tubes with the 

 apparatus at the left of fig. 2, From the large bottle of stock 

 solution it was allowed to run into the tube A until the fine tube 

 in the top of this just broke the meniscus. The escape was then 

 connected by a rubber joint with the small end of the absorption 

 tube, and the liquid allowed to run out until the tube Inserted in 

 the lower end of A broke the meniscus ; this measured out 100^*^ 

 without its coming in contact with any CO^ except in the air in 

 the absorption tube. Of course the stale baryta water standing 

 in the escape of ^ was washed out just before the tube was filled. 

 No baryta water was allowed to stand In the small bulb of the 

 Winckler tubes. 



After each interval of experiment the solution used was 

 poured from the absorption tube into a bottle holding a little 

 over lOO*'*', without exposure to air except that in the bottle; 

 and the bottle was immediately placed in the position of B in 

 fig. -?, or else stoppered. If respiration had been active enough 



