104 



decided on glucose, one part of glucose to throe of water, tfp. Gr. 1.21. In 

 this mixture the light kernels came to the top. I thus had light and heavy 

 kernels, and with these I experimented as follows: 



First Test. — 300 kernels were taken from every other row of an ear 

 that tested "good" in the usual "seed box" germinating test : Lot 1, 300 

 kernels from the alternating rows were divided into two lots by the specific 

 gravity test; lot 2 showed 258 heavy grains; lot 3, 42 light grains. (Lot 

 1 was not put in solution.) The vitality of these three lots was deter- 

 mined by testing in a box, under identical conditions. 



GERMINATING RESULTS. 



Lot 1 (300 grains) 86% germinated 



Lot 2 (heavy kernels) 89% germinated 



Lot 3 (light kernels) 69% germinated 



Second Test. — 100 kernels were taken from an ear which showed a 

 germinating test of 4 dead kernels and 1 weak out of 5 (the usual test 

 number being 5). These kernels were separated by the specific gravity 

 test and tested as before. 



GERMINATING RESULTS. 



Of the 68 heavy kernels 47% germinated 



Of the 32 light kernels 15i% germinated 



Third Test. — 100 kernels were taken from two ears which showed 

 "extra strong" in the germinating test. They were separated by the specific 

 gravity test, which gave a high percentage of heavy kernels, and were 

 tested under conditions similar to the above. (It will be noticed that the 

 per cent of light kernels is quite small and that all germinated.) 



GERMINATING RESULTS. 



Heavy kernels (91 ) 100% germinated 



Light kernels (9) 100% germinated 



Fourth Test. — Two full rows were taken from 25 ears, in which all 

 of the five test kernels had failed to germinate in the "box test." This 

 gave a total of 2,116 kernels. The specific gravity test showed 592 heavy 

 and 1,524 light ; these were tested as mentioned above. 



GERMINATING RESULTS. 



Of the 592 heavy kernels 54% germinated 



Of the 1,524 light kernels 22% germinated 



