430 University of California Publications in Botany [Vol. 5 



• 



it was our intention to carry out an extensive series of experiments on 

 a number of pure lines and hybrids of Nicotiana in which the number 

 of pollen grains involved in each pollination was accurately known. 

 Limitation of time and the lack of adequate greenhouse facilities have 

 made it impossible to carry out our original plans. However, the 

 results which have already been obtained seem of sufficient interest to 

 warrant their publication at this time. 



The particular results which are the subject of this paper were 

 obtained in experiments upon six plants of Nicotiana Langsdorffii var. 

 grandiflora (U. C. B. G. 107/08w). The pollination experiments were 

 carried on in a greenhouse the average temperature of which was 30° C. 

 It was known that the species used did not exhibit any traces of a 

 parthenogenetic nature. 



Flowering laterals were trimmed to a large bud and all leaves were 

 removed. After careful castration these buds were bagged. When 

 fully receptive a pre%aously determined number of pollen grains was 

 applied to the stigma and all pertinent data recorded on the copper 

 wured pot-label which served to attach the bag to the plant. The 

 paraffin bags were removed about two weeks after pollination. 



The number of pollen grains used in the various pollinations was 

 controlled in the following manner. A few grains were placed in the 

 depression of a hanging-drop slide. Their number was determined 

 and the preparation was covered with a cover slip and sealed with 

 water. In the greenhouse the corolla of the flower was turned back, 

 the cover slip removed, the slide carefully turned over and the pollen 

 allowed to come into contact with the stigmatic surface covered with 

 stigmatic secretion. Another count of the pollen grains remaining on 

 the slide was then made to determine the number that had become 

 attached to the stigma. Obviously defective grains were not considered 

 in the pollen counts. The sources of error in any such method are 

 obvious. However, the results given in the following table indicate 

 that this method of obtaining controlled pollinations is approximately 

 accurate. Other methods were tried but proved in every way less 

 satisfactory than the one described above. 



At maturity the ripened seed capsules were gathered and the 

 number of seeds counted. It was possible in a few cases to determine 

 the position of the seeds on the placental surfaces. None of the seed 

 was tested as to its viability, but microscopic examination demonstrated 

 the presence of normal embryos and endosperms. The results of some 

 twenty-one controlled pollinations are given in the following table. 



