DEPARTMENT OF EXPERIMENTAL EVOLUTION. 8l 



chosen parents proved to be heterozygous. A wild strain, known in Germany 

 as Melandrium riihrum, has been found homozygous for the red color ; but 

 this form is not wholly satisfactory for experimental purposes, because of 

 its fixed biennial habit. The effort to secure a homozygous red-flowered 

 strain among the annual forms has therefore been continued. New features 

 in the color-inheritance of Lychnis are the discovery of a "dominant white," 

 which has given an F^ mainly white-flowered when crossed with a homozy- 

 gous red-flowered individual, and an all purple-flowered F^ when crossed 

 with a recessive white. A new Mendelian unit-character has been demon- 

 strated by the unexpected appearance of a type of Lychnis having yellowish- 

 green foliage, which is hypostatic to the full dark green of the usual type. 

 Several sorts of variegation and several bud-sports are being investigated. 



Poppies. — The problem of analyzing the hereditary factors which serve to 

 differentiate the many garden derivatives of Papaver rhocas from each other 

 and from their wild prototype has been resumed. The orange-red body- 

 color of the petals in the wild form is epistatic to all the shades and tints 

 displayed by the garden forms, but the presence of a distinct margin and of 

 doubling often found in the cultivated strains are epistatic to the self-colored, 

 single form of the wild poppy. Interesting phenomena of "repulsion" and 

 inhibition of characters have been discovered and appropriate crosses have 

 been made for further study. 



AN ABERRANT INHERITANCE RATIO AND ITS INTERPRETATION. 



On the principle of segregation of the determiners of characters, the char- 

 acters should appear in the offspring of hybrids in certain definite ratios. 

 An exceptional case appeared in breeding two kinds of shepherd's purse in 

 the ratio of the two forms of seed-capsules and some leaf-characters that re- 

 appeared in the second hybrid generation. After several years of work Dr. 

 ShuU has reached a nearly complete explanation of the aberrant ratios, as 

 follows: (a) There are two independent genes determining the triangular 

 capsule, thus leading to the expected ratio 15:1 in the F2 and both 15:1 and 

 3 : I in the F3 ; (&) the heegeri individuals are less successful in the embry- 

 onic stages than the bursa-pastoris individuals, so that the ratio 15: i be- 

 comes about 22: I, and 3: i becomes 4.67: i ; (c) that there is a variation in 

 the relative potency of the genes determining certain features of the rosette- 

 leaves which has led to the classification of a portion of the heterozygotes 

 with the negative homozygotes. Further data are being sought bearing on 

 these points. 



THE INJURIOUS EFFECTS OF SELF-FERTILIZATION OF CORN. 



Dr. Shull has continued his researches, which have revealed the great 

 complexity of the population of any cornfield. Self-fertilization, as is well 

 known, yields seed that develop into inferior individuals, while outcrossing 

 usually gives vigor. Dr. Shull has been working on the hypothesis that the 



6 — YB 



