36 BURSA BURSA-PASTORIS AND BURSA HEEGERI : 



hybrid between B. bp. Itclcris and />'. bp. rhoinboidea. Had B. bp. helcris 

 been a normal component of this family it would have made up a large 

 portion of it, instead of less than 1 per cent. As none of the earlier cul- 

 tures were guarded against cross-pollination, it seemed more reasonable 

 to consider this one specimen the result of a chance cross produced by 

 pollen from 77. bp. Jictcris, which was abundantly represented in the house 

 at the same time. To test further the character of the normal components 

 of this family as well as of this one atypic individual, the following two 

 ciiltures were made : 



05190.167: A typical 7?. bf>. rhomboidea, the type of the last family, 

 whose seeds were sown July 4, 1906, grave 29 plants, all B. bf>. rhoinboidea, 

 showing- the expected purity of this extracted form. The pollination of 

 this plant had been carefully guarded. 



05190.178: This was the atypic plant described under the last family 

 but one. The pollination was guarded, and the seeds sown July 11, 1906, 

 resulted in a progeny which contained 29 B. bp. heteris, 29 B. bp. rhoin- 

 boidea, and 37 intermediate between the two. The deficiency of the inter- 

 mediate or heterozygous class is again evidence of the difficulty of distin- 

 guishing bet\veen the extreme variations of these and either pure-bred 

 parent. The result here secured is sufficient demonstration of the simple 

 hybrid character assumed for the parent. 



054.208 : This plant had the attenuate primary lobes rather distant and 

 entirely lacked rounded secondary lobes. The sinuses of the earlier climax 

 leaves did not nearly reach the midrib, but later leaves \vere cut more 

 deeply. Seeds of this typical B. bp. tennis were sown March 2, 1906, and 

 gave a progeny consisting of 24 B. bp. tennis and 1 B. bp, heteris. As B. 

 bp. heteris is dominant over B. bp. tennis, it seems likely that the specimen 

 of the former biotype in this family represents another instance of chance 

 crossing, since the pollination was unguarded, as in the other early cul- 

 tures. If this assumption is true, this family represents another instance 

 in which an extracted form breeds true. 



054.209 : In the climax leaves this plant had the typical form of B. bp. 

 heteris. Later leaves of the rosette showed a nearly or quite complete sup- 

 pression of the rounded secondary lobe. Seeds were sown March 2, 1906, 

 but very few plants were secured. These were not well studied, but we re- 

 observed to be heterogeneous, though most were related to B. bp. heteris, 

 the parent form. 



054.210 : This plant was a marked example of B. bp. tennis, remarkable 

 for the extreme attenuation of its lobes. The seeds were sown March 2, 

 1906, and of the 31 plants capable of classification, 27 were B. bp. tennis of 

 strongly marked type, and 4 were B. bp. heteris. Unless these 4 were 



