i;xpe;rime;nts in bre;e;ding sweet corn. 285 



the actual case) the improvement in earhness is retained there 

 in successive generations it is quite impossible to be certain 

 whether this is because it is inherited or because the corn is 

 quite perfectly adjusted to the local conditions. In all breed- 

 ing work with corn certainly, and probably with many, if not 

 all, other crops as well, these two factors must be very carefully 

 distinguished, or incorrect conclusions are almost certain to be 

 drawn. 



It, therefore, seemed desirable to test on a large scale what this 

 selected seed would do under a wide variety of environmental 

 conditions. By cooperation with the packers mentioned above 

 (p. 255) a farm distribution test was made in 1909. Through 

 their aid it was possible to get from one-half to 2 acres of corn 

 from this seed grown by each of 47 farmers. The name and 

 address of each of these farmers together with the acreage 

 grown are given in Table 9. 



It should be said that the acreages given in this table are not 

 estimates, but were determined by actual accurate measurement 

 of the plots. Taking all the cooperating planters together the 

 average acreage in the test per planter was 0.95 acre, an insig- 

 nificant fraction under one acre. The largest number (21)- 

 planted between i and i^ acres. The nejct largest number 

 (17) planted from ^4 acre to i acre. Only 6 out of the 47 

 planted less than y^ acre, and only 3 more than i^/^ acres. Of 

 the latter only one (Mr. Heath) planted more than 2 acres. 



The largest number of growers in any single county was in 

 Oxford (15). Cumberland and Androscoggin counties each 

 had 6 ; Franklin and Kennebec each 5 ; Penobscot 3 ; York, Som- 

 erset and Waldo each 2; and Knox i. This distribution fairly 

 well covered the corn growing region of the State (cf. fig. 231), 

 and included a wide range of environmental conditions. 



