EXPERIMENTS WITH TOBACCO BLOB80M8. 11 



used iii the field, though only two months old, was exposed to, and 

 perhaps injured by, the hot weather of July and August, while thai 

 used in the greenhouse experienced the cooler weather of April. May. 

 and June. When fresh pollen was used the destructive effect of pre- 

 mature pollination was as pronounced on the flowers in open air as on 

 those in the greenhouse. That a saturated atmosphere is sufficienl to 

 start the growth of tobacco pollen was shown by placing some on the 

 under surface <>t* a rover glass over a drop of water in the well of a 

 microscopic Blide. The next day pollen tubes were found growing 

 from many of the grains. 



The seeds resulting from these experiments were not all planted to 

 test their germination, but seeds from several of the tobacco experi- 

 ments were tested and the large plump ones (PI. III. fig. 6), which, 

 when crushed, Bhowed that they contained oil. germinated well, while 

 the smaller, light, chaffy ones failed to germinate. It was therefore 

 thought unnecessary to test the germination of all the seed, and good 

 ones were distinguished for the most part by their appearance. 



When the first tlowers were operated upon— i. e., those of experi- 

 ments 1. -1. and 3— very little attention was paid to the age of the 

 flowers chosen, other than to see that they had not yet opened. ^ This 

 is now known to account for the great variation in the number of pods 

 obtained. (See Table I.) As the work proceeded, however, it was 

 found that the age of the flower at the time of pollination was of 

 the utmost importance in order to obtain results of value, and the time 

 was recorded in hours in performing all experiments after the first 

 six. Finding that the age of the flower operated upon was the most 

 vital condition entering into the experiments, a very careful study of 

 the growth of the flowers was made. A number of young flower buds 

 of various ages were labeled with tags upon which were placed the 

 date and size of the bud, and each day thereafter the size and condition 

 of these buds and the resulting flowers and seed pods were written on 

 the respective labels. This was repeated several times. The weather 

 conditions were found to exert a marked effect on the rapidity of 

 development, but it was noticed that the size and appearance of the 

 corollas were good indications of the advancement of the inclosed 

 anthers and stigmas, so that consequently the corollas were used in 

 the experiments as a means of designating the age of the flower buds 

 operated upon. The 35 experiments performed with tobacco to settle 

 the destructive influence of premature pollination are outlined and 

 their results given in Table 1 (p. 17). Tobacco being a self-fertilizing 

 flower, the stigma is receptive at the time the anthers open, and the 

 anthers usually begin to shed pollen when the flower begins to open. 

 Comparing the length of the corolla with the length of time before 

 the corolla is fully open and the stigma receptive, we find the following 

 conditions to exist when the plants are thrifty and the weather fairly 

 favorable: 



