No. 556] DEVELOPMENT IN NICOTIAN A 215 



In the cultures at the Bussey Institution of Harvard 

 University the normal Cuban is known as 402 : the fasci- 

 ated strain as 300-309, the range in numbers represent- 

 ing an attempt at selection. 



Methods 



Data were collected on each plant and tabulated sepa- 

 rately. The characters noted were, extent of fasciation 

 in the stem; number of leaves, petals, sepals, stamens 

 and ovary locules. Twenty-five flowers were taken from 

 each plant and the parts of each flower recorded sepa- 

 rately. In all crosses made the flowers were castrated 

 in the bud and bagged. Pollen was taken only from 

 anthers still in the closed bud and 95 per cent, alcohol 

 was liberally used after each operation on hands and 

 instruments. The Webber system of recording the 

 plants by number was used. All seed was sown in steril- 

 ized soil and all possible care taken to avoid mixtures. 



Experimental Work 

 Numerous crosses were made between distinct species 

 and the abnormal race, but all of the progeny were ster- 

 ile, though the abnormal character was visible in their 

 flowers and in the increased number of leaves which they 

 bore. 



Four crosses were made between normal N. tabacum 

 varieties and the abnormal, all of which produced fer- 

 tile F x plants. The most interesting of these is a cross 

 of the abnormal with the normal Cuban variety from 

 which it mutated. Three generations of this cross 

 (304 x 402) have been grown. The F 1 generation con- 

 sisted of 39 plants, the F 2 of 97 and the F 3 of 647, total- 

 ing 783 individuals. The F, was intermediate in char- 

 acter between the two parents, as the table will show. 

 The F 2 gave the three expected types in the ratio of 

 1:2:1, the actual numbers being: 



