12 INJURIOUS EFFECTS OF PREMATURE POLLINATION. 



Four days before the stigma is recepth e the corolla does not extend 

 beyond the calyx lobes, which are one-half inch long-. The flower 

 labeled c, in PL I, fig. 1, is in this stage of development and is the 

 same as those which were chosen for experiments 22 and 35. (See 

 Table I.) The growth of an unmolested bud of this age can be studied 

 by comparing the small unlabeled bud on the right in PI. I, fig. 1, with 

 resulting flower and seed pod in PI. I, fig. 2, and PI. II, fig. 1, 

 respectively. 



Three days before the stigma is receptive the corolla is one-fourth 

 to one-half inch longer than the calyx (flowers d and 6, PI. I, fig. 1). 

 Flowers of this age were chosen for 10 of the experiments given in 

 the table. 



Two days before opening, the corolla is three-fourths to 1 inch 

 longer than the calyx, as shown by flowers 1 and 2 in PI. I, fig. 1. 

 Flowers of this age, as well as younger ones, were found to be killed 

 by pollination. (Compare figs. 1 and 2, PI. I.) 



One day- before opening, the corolla is 1\ to If inches longer than 

 the calyx, the terminal one-fourth inch being of a red color. Flowers 

 3, 4, 5, and a, PI. I, fig. 1, are of this age and can be pollinated with 

 fair success (experiments 1, 6, etc.), but the results will be much 

 better if the work is done twelve hours later (as in experiments 16 

 and 36), or if the flowers are emasculated at this age, or younger, and 

 pollinated when the pistils become receptive (experiment 7 or 20). 



The next day the corolla and anthers open at the same time and the 

 receptive pistil is of proper length and maturity to receive the pollen 

 directly from the opening anthers. One month after pollination the 

 seed pod is ripe. 



Having obtained this data in regard to the manner of growth, the 

 flowers were designated in the notes made at the time the experiments 

 were performed as " three days before pistils would have been recep- 

 tive,'" "'one day before pistils would have been receptive," etc. 



By the time the results of the first 10 experiments were obtained it 

 was definitely known that an abundance of fresh pollen applied to 

 tobacco flowers two or three days before the pistils become receptive 

 results in the falling of the flowers without setting fruit, while flowers 

 operated upon one day before becoming fully open and receptive give 

 fair percentages of fruits, as shown by experiments such as Nos. 7 or 

 8. But it was necessary to find out why the former failed while the 

 latter gave fair percentages. It might be supposed that the younger 

 flowers were more injured by emasculation, but several experiments 

 soon proved that this is not the case. In experiment 20 the flowers 

 were emasculated three days before becoming receptive and pollinated 

 when receptive, resulting in 100 per cent of fruits, while in experi- 

 ment 21 they were emasculated and pollinated three days before 

 becoming receptive, resulting in no fruit whatever. The only differ- 

 ence in experiments 20 and 21 is in the time of pollination. Other 



