EFFECT OF MOISTURE AND TEMPERATURE. 31 



constructed that the. relative humidity of the inclosed air could be 

 inn-eased without the admission of more free air. The detailed con- 

 struction of this apparatus is shown in fig. -2." 



The seeds continued to absorb moisture to a limited extent. In order 

 that the inclosed air might be maintained at approximately the same 

 degree of saturation, a crude hygroscope was attached on the inside of 

 eacb bottle. These h} T groscopes were made from awns of Stipa 

 capillata L., the tip of the awns being removed and a short piece of fine 

 copper wire used as an indicator. These hygroscopes were suspended 

 from the under side of the cork, as shown at A, and by the side of each 

 was suspended a fine fiber of silk, which, being carried around by the 

 indicator, recorded the number of turns made by the awn. 



Five such preparations were made for each of the two sets, so as to 

 furnish seeds for a series of tests. One set was kept at a temperature 

 of 30 to 32 C., the other at 36 to 37 C. The seed from one of the 

 bottles, at each of the temperatures, was weighed after eighty-one 

 days, at the time the germination tests were made. These weighings 

 showed that at the lower temperatures the average increase in weight 

 for all the seeds was 8.6 per cent, and at the higher temperatures, 6.3 

 per cent. The increase in the case of the beans was quite marked at 

 this time, being 13.3 per cent for those maintained at a temperature 

 ranging from 30 to 32 C., and 9.8 per cent for those maintained at 

 36 to 37 C. 



The third set of conditions consisted simply of packages of the air- 

 dried seeds kept in open boxes in each of the incubators. This series 

 of tests was made especially for the purpose of determining the effect 

 of dry heat on the vitality of seeds when maintained at the tempera- 

 tures above given for some considerable time. 



For the fourth series small packages of the seeds were put into 

 2-ounce bottles, which were then corked and sealed with paraffin. Five 

 of these bottles were kept in each of the ovens and germination tests 

 were made at irregular intervals. The results of these tests furnish a 



The wide-mouth bottle (b) contains the packages of seed (s). Through an open- 

 ing in the cork is inserted a short piece of soft glass tubing, being first fused at the 

 lower end and having a slight constriction drawn at c. At a distance of 1 cm. 

 above the constriction is blown a small opening, as shown at o. A short piece of 

 heavy rubber tubing (0, cemented on a piece of heavy brass wire (10), serves as a 

 stopper. This stopper, which must fit closely within the glass tube, is operated by 

 means of the heavy wire. When drawn up, the water in the tube may give off 

 aqueous vapor, which can escape through the small opening (o) into the bottle. 

 When sufficient moisture is present the supply is shut off by pushing the stopper 

 down firmly against the constriction. The stopper must be well coated with vas- 

 eline to prevent its sticking to the sides of the glass tube. To make the apparatus 

 more secure against the entrance of fresh air, a second piece of rubber tubing (r) 

 is placed in the upper part of the glass tube, the top of which is then filled with oil. 



