♦ Sorting and blending of unit cliaracters. 203 



to above where seed was sown in October 1912, on ground recently 

 turned. In such cases no autumnal rosettes are formed. Since the 

 rosette leaves developed in late summer and early autumn present a 

 higher degree of character differentiation and complex than the spring 

 and early summer leaves, the complete life cycle and full expression of 

 the species is not obtained. The annuals of biennial species reach the 

 flowering and fruit period by short-circuiting the complete life history. 

 They are really short-circuit forms. Sometimes in green house cultures 

 seeds falling from ripe pods of a potted plant during the winter, ger- 

 minate, and under the warm conditions of hot house culture begin stem 

 development early, omitting the rosette stage. Such short-circuit forms 

 in species where late summer and early autumn leaves of autumnal 

 rosettes are strongly cut over the basal portion, and the stem leaves 

 are only slightly cut, or toothed, suggest a variation in the leaves due 

 to the influence of the changed environment, and thus may be mis- 

 leading. The cycle has been shortened from early rosette leaves slightly 

 cut to lower stem leaves slightly cut, omitting the late formed strongly 

 cut leaves of the mature rosette stage ^). All gradations sometimes appear 

 between the extreme short-circuit forms and those with the rosette stage 

 complete, the degree of rosette development reached depending on the 

 time at which stem development begins. So far as I am aware the 

 conditions have not yet been analyzed which determine the time of stem 

 development in these annual forms of biennial species. 



In connection with this reference to annual forms, it may be of 

 interest to note the tendency to a perennial condition which appears 

 in Oe. nutans. I have briefly refered to this tendency in another 

 place ''^). Plants which had matured and gone to seed in late August 

 and early September, put out new shoots with flowers in late Septem- 

 ber and early October, from the old bare stems. A few of these were 

 potted and taken into the green house where the}' continued to 

 flower all winter. A few of those in the garden, also in the green 

 house, formed autumnal rosettes at the base of the stem. Those in the 

 garden were put into large pots, one was plunged in the soil of the 



^) Exactly such short-circuit annual forms appeared iu an undescribed species, 

 No. 17 of my 1913 cultures, about 50% being short-circuit forms, while the remainder 

 formed mature rosettes with leaves strongly cut over the basal half, which now (1914) 

 are coming into flower. 



-) Atkinson, Geo. F. Is the biennial habit of Oenotheraraces constant in 

 their native localities? Science N. S. 37, 716, 717, 1918. 



