The Isolation of Tricofyloiis hitcniicdiatc Races. 431 



among them all, the plant with 16% would, no doubt, 

 have offered me as good a prospect as the corresponding 

 plant of Clarkia. The chosen plant was remarkable for 

 the large proportion of tetracotyls and the low proportion 

 of hemi-tricotyls which it produced. There were 1 1 % 

 of the former, and only 1 % of the 

 latter. This peculiarity has reappeared 

 amongst its descendants, especially with 

 regard to the tetracotyls. 



In 1895 I planted out on separate 

 beds tricotyls and tetracotyls of the 

 parent with 41% only. Among the 

 tricotyls ^2 plants set plenty of seed. 

 Of these one produced only 6% tri- 

 cotyls, amongst the rest the values 

 were distributed between 13 and 43%, 

 and their mean was at 26%. The 

 tetracotyls gave similar numbers, em- 

 bracing 19 plants with from 14 to 42% 

 and a mean of 25% ; besides these there 

 was one plant with as much as 51%. 

 This latter fact must ob\iously be at- 

 tributed to a fortunate chance, and we 

 may conclude that the tetracotyls are 

 not more likely, nor on tlie other hand 

 less so, to produce tricotyls tlian the 

 tricotyls themselves, but that they ob- 

 viously belong to the same race, i. e., 

 tliat their character is brought about bv the same elc- 

 mentary factor. Tlie proportion of tetracotylous seed- 

 lings in this culture was very high, but not higher tlian 

 the ratio recorded for the first generation. 



For tlie continuation of the race onlv seedlings of the 



Fig. 86. HcUchry- 

 suni bractcatuni 

 conif>osifu}n. A 

 flowering stem of 

 a plant of the tri- 

 cotylous interme- 

 diate race. 



