1908] SHULLMENDELIAN INHERITANCE HI 



From all of these results it must be concluded that the purple disk 

 is a strict Mendelian character, and that the first purple-disked 

 parent was a heterozygote. The original cross is being repeated to 

 make the latter point more certain. 



LYCHNIS DIOICA L. 



This plant is a native of Europe and was doubtless very often 

 cultivated in the gardens of a generation or more ago. It has now 

 commonly escaped from the gardens and is maintaining itself quite 

 generally along roadsides in the eastern United States. The flowers 

 show considerable variation in the intensity of their purple color, and 

 white-flowered ones are quite common. Purple and white flowers 

 have been crossed in so many species and found to follow MENDEL'S 

 law, that there would have been little incentive to study this particular 

 character in Lychnis dioica, had it not been for the fact that this 

 species is dioecious, and that in consequence self-fertilization could 

 not be held responsible for the Mendelian behavior if it should be 

 found to exist. Statements in the manuals regarding the dioecious 

 habit of this species indicate that it also occurs as a hermaphrodite, 

 but among about five thousand individuals which I have examined 

 only one was noted which had both stamens and pistils, and that one 

 had but a single flower with both kinds of sporangia. 



During the summer of 1905 I made all possible combinations with 

 white and purple flowers, but in one case unfortunately used two 

 different purple-flowered individuals for the pollination of the same 

 white-flowered plant and collected the seeds from both crosses together. 

 The offspring from these crosses reared in 1906 were as follows: 



No. 05212, w X w, 90 w 



No. 05213, w X 2 p's, 66 p:4i w 

 No. 05114, p X p, 65 p:23 w 

 No. 05115, p X p, 96 p 



The fact that two white-flowered parents produced nothing but 

 white-flowered offspring, and that in one case two purple-flowered 

 parents produced only purple -flowered young, shows pretty conclu- 

 sively that we are dealing here with a Mendelian pair of characters, 

 even though the one family (05114), which probably represents the 

 cross DRxDR, shows a considerable departure from the expected 



