Hayes : Variation in Tobacco 



43 



seed was saved from the two most 

 desirable plants, which bore 19 and 20 

 leaves respectively. As a measure of 

 variability the number of leaves per 

 plant was noted and observations on 

 general plant habit were also made. 

 The two sclfed plants were given the 

 selection numbers 13-2 and 13-1, and 

 their progeny was grown in 1911. 



The variability was also large in the 

 second season, 13-1 being as variable 

 for leaf number as in the first year, but 

 there was a much larger percentage of 

 plants with good-sized leaves than in 

 the original lot. In 13-2 there was a 

 much larger percentage of desirable 

 plants. Two plants of strain 13-2, 

 bearing 18 and 19 leaves respectively, 

 were self-fertilized and grown the fol- 

 lowing season. The plants in this third 

 season proved very uniform both in the 

 number and shape of leaves. As far as 

 field characters go, these selections were 

 the equal of the commercial Cuban 

 tobaccos grown on the same field. The 

 quality of the cured leaf was also very 

 good, but no better than the' present 

 Connecticut Cuban shade variety. 



Much variation was shown in the 

 statistical results for number of leaves 

 per plant. It is this variation that has 

 caused the belief in the efTicacy of en- 

 vironment as a means of inducing a 

 breaking up in type. More recent 

 knowledge of heredity convinces us 

 that the apprent breaking up of type is 

 simply an expression of the mechanical 

 and germinal mixture of characters. 

 Selection and inbreeding tend to pro- 

 duce homozygous forms, and the length 

 of time necessary to obtain pure forms 

 is largely dependent on the number of 

 row cultures which may be grown, 



careful test made. 



To determine whether the variations 

 observed in generations of the above ex- 

 periment had a connection with change 

 of environment, three seed pods picked 

 from different plants growing in Cuba 

 were carefully har\^ested and sent in sep- 

 arate compartments of a package from 

 Cuba to ]\Ir. Stewart. The seed from 

 each pod was sown in sterilized soil, the 

 seedlings being marked respectively 

 CI, C2 and C3. A row of each of the 



selections was grown on the same land 

 as the selections 13-2-3 and 13-2-4 

 previously mentioned. 



Each of the row cultures had a very 

 uniform habit of growth. CI and C3 

 grew to about the same height, but the 

 average size of leaves of C3 was some- 

 what larger than for CI. C2 had about 

 the same average size of leaf as C3. 



An examination of the results shows 

 that CI and C3 both produced about 

 21.3 leaves per plant and were not very 

 variable for leaf nimiber. It is possible 

 that one 13-leaved plant of C3 really 

 belonged to C2 and was accidentally 

 mixed when sown. C2 only gave an 

 a\'erage mean of 13.83 leaves per plant 

 and a range of variation from 11 to 18 

 leaves. These selections show about 

 the same variability, as determined by 

 S. D. and C. V., as 13-2-3, 13-2-4 and 

 the commercial Cuban, and indicate 

 that there has been no breaking up of 

 type due to changed environment. 



In connection with a study of in- 

 heritance of separate characters of 

 tobacco, some observations have been 

 made on the effect of changed environ- 

 mental conditions. While all quanti- 

 tative characters are greatly dependent 

 on environment for their full de\'elop- 

 ment, some characters are more notice- 

 ably affected than others. The most 

 uniform character of all was found to be 

 the number of leaves per plant. Each 

 of four selections was grown both at 

 Forest Hills, Mass., Bloomfield, Conn., 

 and New Haven, Conn., from seeds of 

 a single plant. 



In order to have the counts as uniform 

 as possible and yet allow the crop to be 

 harvested, the following method was 

 used. The number of leaves per plant 

 was counted from the fourth leaf from 

 the bottom to the leaf below the bald 

 sucker at the top, "bald sucker" being 

 the farm name for the first sucker which 

 appears without true leaves. This 

 method allowed the selections to be 

 topped and also gave added interest, as 

 the leaf counts represented about the 

 average nimiber of leaves which were 

 harvested. 



The field at Forest Hills was very 

 fertile but in a region where tobacco 



