263 





d 



^ij_tei 



pericarp is -equivalent to its removal. In either case the dormant 

 grains are killed within a comparatively short time (75 days). 



It will be noticed that grains with the pericarp removed but not 

 otherwise treated (Table 10) retained the power of germination to 

 a fairly high degree for the time indicated in the table ; but such grains 

 after ten days' exposure to the dry air of a steam- 

 heated room produced seedlings that were uni- 

 formly weak. This indicates that these mem- 

 branes play a very important role in preserving 

 the vitality of the grains. 



The dry pericarp, not including the pedicle, 

 was shown to be impermeable to kerosene in an- 

 other way. A large grain of the Champion White 

 Pearl furnishes a membrane fully one-half inch 

 in diameter. It is easily removed after soaking 

 the grain for a few minutes in warm water. 

 After drying, it can be cemented over the end of 

 a glass tube for use either as a barometer or as 

 an osmometer. This simple piece of apparatus 

 was, as far as I know, first devised by Becquerel 

 ('07) in his studies on the permeability of seed 

 coats to certain gases. Shull ('13) also used it 

 with success in demonstrating the semi-permea- 

 bility of the testa of Xanthium. Adapted for my 

 work, the apparatus was constructed as shown in 

 the accompanying sketch. 

 Considerable difficulty was experienced in finding a cement not 

 soluble in kerosene. Sealing-wax, such as is used by express com- 

 panies for sealing valuable packages, was finally found to serve the 

 purpose admirably. The rubber stopper at the end of the glass cylin- 

 der serves as a foundation to which the m-embrane is cemented. The 

 small glass tube (Fig. i, cc) was allowed to protrude three or four 

 millimeters through the perforated rubber stopper. A layer of wax 

 equal in thickness to th-e protruding portion of the tube was then ap- 

 plied and the edges of the membrane were pressed into it while it was 

 still soft. More wax was then applied to make the seal perfect. The 

 central portion of the membrane over the end of the small glass tube 

 was left entirely free from wax. 



Such an apparatus was s>et up March 7, with kerosene on one side 

 of the membrane and with plaster of Paris as an absorbent on the 

 other. At the present time (May 6) the wax is holding perfectly and 

 there has been no trace of oil passed through the membrane. A 



Fig-. 1. Apparatus used 

 in testing directly the per- 

 meability of the pericarp 

 of Zi-a: a, membrane; bb, 

 rubber stopper; cc, glass 

 tube; i/, receptacle for 

 ti<)uid to be tested. 



