438 EXPERIMENTS WITE PLANTS 



nearly 7 per cent iii six years. The method at first 

 involved a rigid chemical analysis of the ears and 

 selection of the best. Afterward it was found that 

 for jiractieal pm'poses the chemical analysis could be 

 dispensed with: since the germ contains most of the 

 oil, it is only necessary to select the kernels with 

 the largest germs. On the other hand, Corn with 

 a lower oil-content is wanted as a feed for hogs, since 

 it produces harder, firmer bacon. So the Station pro- 

 ceeded by selection to decrease the amount of oil 

 to less than "1 per cent. It would, therefore, depend 

 on market conditions whether we should try to increase 

 or decrease the amount of oil in the corn. 



In the same way we should breed to increase the 

 amount of protein if the corn is to be used for food, 

 but to decrease it if the corn is to be used for starch. 

 The Station workers were able to increase the pro- 

 tein from 10.92 per cent to IG per cent in about six 

 years, and to decrease it to 6.66 per cent in the same 

 Lnigth of time. Here, again, it is found practicable 

 for ordinary purposes to dispense with chemical 

 analysis and select those kernels which have the white, 

 starchy part around the germ best or least developed, 

 according to which is desired: this is possible be- 

 cause the protein is almost all contained in the germ 

 and in the horny outer part of the kernel, while 

 the starch is practically all in the white portion which 

 lii^s between them. 



