Apr. 1894.] BOTANICAL SOCIETY OF EDINBURGH. 197 



A duplicate to this piece of apparatus, similar iu every 

 respect, was prepared, but light was excluded from its 

 interior by an external coating of very thin sheet brass. 

 The two tubes were fixed in an appropriate holder, and the 

 whole was immersed under the surface of water, with only 

 the open upper ends of the graduated tubes and the water 

 joints protruding. The trough itself was placed in strong 

 diffused light in a small greenhouse facing the north. 



o o o 



Both pieces of apparatus were carefully calibrated. 



Method of Experiment. — A sample of the corn, about 

 50 grams of barley or wheat, was taken and placed in a 

 beaker. The corn was covered with water from thirty-six 

 to seventy -five hours, changing the water at intervals. 

 The water was then drained off the seeds, and they were 

 placed in a glazed earthenware jar, covered with an opaque 

 plate, and allowed to germinate for periods varying from 

 two to eight days. At the stage fixed on for experiment 

 the jar was opened and the sample turned out. Thirty 

 corns, or thereabout, were then selected for each tube, the 

 greatest pains being taken to ensure similarity in the size, 

 weight, and progress in germination of the two samples. 

 It was hoped by employing this method of selecting the 

 corns to eliminate in some degree the variations caused by 

 differences in the samples employed for experiment. After 

 being weighed, the samples were introduced into their 

 separate tubes, as shown in the figure, and the apparatus 

 was immersed. After a short interval, to establish equili- 

 brium of temperature, readings were taken on the limbs of 

 the tube D, which is half filled with cj^uicksilver, and the 

 volume of the enclosed air was deduced after allowing for 

 temperature, pressure, tension of aqueous vapour, and the 

 volume of the seeds, as well as that of the two boats with 

 their liquid contents. The volume occupied by the corns 

 was ascertained by finding the specific gravity of a separate 

 portion of the same sample. The readings were repeated 

 at suitable intervals. The reduction in volume between 

 each observation gave the amount of oxygen absorbed by 

 the seeds, the carbonic anhydride produced being absorbed, 

 as soon as formed, by the potassa solution in the boat B^. 

 "When it was required to change the air in the tubes, which 

 was usually done immediately after an observation of the 



