259 



Champion White Pearl was removed from the kerosene in which it 

 had been placed February 6 — 76 days' immersion — and 50 grains were 

 selected from it which seemed from external appearances to be free 

 from kerosene. These grains were exposed to the air for ten days 

 and then planted in a 30% saturated soil. Forty-nine, or 98% of 

 them, developed normal seedlings. From these results it is apparent 

 that an almost perfect germination can be secured by selecting grains 

 showing no traces of stain in the embryos. The selection of grains 

 with membranes slightly permeable to the colored oil can by no means 

 so easily be made. The structures at the tips of the grains always take 

 up the oil readily and it spreads for some distance from them, giving 

 the appearance of stain within the embryos when in reality it is en- 

 tirely superficial. The absence of Sudan III in the grain does not 

 necessarily mean the absence of kerosene, since the membranes may 

 be semi-permeable. A number of experiments were undertaken to 

 test this assumption. 



Grains immersed in Sudan III kerosene for long periods and free 

 from stain were carefully dissected, and the structures within the 

 coats were tested by the picric acid methods of Schulz ('08), and 

 Krauz ('09). The results were uniformly negative. It was found 

 that tests by these reagents were not nearly so delicate as the sense 

 of taste. In no case, however, could the presence of kerosene be de- 

 tected in unstained grains. On the other hand, it could be readily de- 

 tected in grains which had been but slightly stained with Sudan III. 

 It should be here stated that the above holds true only for air-dry 

 grains. 



These facts indicate very clearly that the kerosene enters some 

 of the grains and is excluded from others. Whether the membranes 

 of the grains showing penetration had been mechanically injured or 

 were of different physical structure has not been determined. In 

 either case the result would be the same. Undoubtedly there are many 

 opportunities for mechanical injuries, but the fact that the number 

 of grains exhibiting a penetration of the kerosene increases with the 

 time of immersion would indicate that the membranes are not uni- 

 formly impermeable. 



Membranes of widely different properties are not uncommon in 

 seeds of the same kind. Many cases of delayed germination are at- 

 tributed to this peculiarity [see Crocker ('06) ; Hiinlein ('80) ; Nobbe 

 and Hanlein ('7?) ]• It is not unlikely that the membranes of the corn 

 kernel are sufficiently different in their organization or development to 

 permit a rather wide variation in their permeability to kerosene. 



A number of interesting studies on the physical properties of 

 plant membranes have appeared recently. Brown ('07 and '09) found 



