432 BROOKLYN BOTANIC GARDEN MEMOIRS 



Several of these ratios deviate widely from the 3 : i ratio, but in 

 each such case the inclusion in the monomeric group has been based 

 on some special consideration. In some cases one or more pedigrees 

 from the same parentage gave a convincing approximation to the 3 : i 

 ratio when grown under favorable conditions; in other cases small 

 samples of the families have been grown under good conditions, and 

 the inclusion of the particular pedigrees in one or another of the tables 

 has been based upon the constitution of these small well-grown samples 

 regardless of the indecisive ratios displayed by the family as a whole, 

 when grown under conditions which tended to suppress the dominant 

 leaf characters here under discussion. In still other cases a number 

 of families belonging to later generations have been grown and have 

 given full confirmation of the classification of the wild biotype from 

 which the pedigree in question originated. Because of the pre- 

 liminary character of the present report, it is not considered neces- 

 sary to present in greater detail, the evidences in support of the 

 conclusion that the families included in this table have a single B 

 factor. It need only be stated that families whose records are inde- 

 cisive for the particular point at issue, have been included for the 

 sake of completeness, and to avoid the immorality of arbitrarily 

 selecting for presentation those cases which are deemed to support 

 convincingly the author's hypothesis. 



Seeds of a specimen of shepherd's-purse, received March 31, 

 191 1, from Tucson, Arizona, through the kindness of Dr. D. T. 

 MacDougal have yielded a pedigree line which has given me much 

 difficulty in the classification of the rosettes, even under the most 

 favorable environment I could provide, owing to the fact that in this 

 particular strain there is so strong a tendency to precocious develop- 

 ment of the stems that the leaf characters even in the climax leaves, 

 are frequently of the relatively undifferentiated juvenile type. It 

 was just in this difficult material that, during several years, two facts 

 impressed me with the probability that there were present in this 

 strain two independent factors corresponding with the B factor of 

 the above notation. These facts were (a) the occurrence of a rela- 

 tively small number of tenuis {Ab) plants in two F2 families (11413, 

 11414) derived from a cross between the Tucson biotype and a speci- 

 men of B. bp. tenuis from the eastern United States. No exact 

 count of the tenuis rosettes in these F2 families was made, but it 

 was noted that these tenuis rosettes were sufficiently distinct from the 

 rest of the family, that their number could probably have been deter- 

 mined with small degree of error. Only desultory attention was being 

 given at that time, however, to rosette characters, owing to seemingly 

 insurmountable difficulties of classification in this biotype, and to my 



