196 Atkinson. 



We are so accustomed to think in terms of Mendelian segregation, 

 i. e. the segregation of distinct hybrid types in the second generation, 

 that the fundamental distinction between the two kinds of segregation 

 may not be fully appreciated at first. According to the Mendelian 

 theory the segregation of the "unit characters" is generally believed to 

 occur during the reduction divisions in the gonotokonts (mother cells 

 of the pollen grains and embryo -sacs in seed plants) at the close of 

 the Fl generation, so that the segregation of the hybrid types occurs 

 first in the second generation. In the production of "twins", "tri- 

 plets", etc., the segregation of "unit characters" occurs in the fertilized 

 egg, or zygote, of the first generation, i. e. at the beginning of the first 

 generation instead of at the close as in Mendelian segregation. 



The segregation of distinct hybrid 'types in the second generation 

 of a cross, according to the number of contrasting allelomorphs in the 

 parents gives rise to "dihybrids", "trihybrids", etc. There is a tendency 

 to confuse dihybrids with twin hybrids, trihybrids with triplets, etc., 

 and it often requires some especial emphasis for the mind to grasp how 

 fundamentally different twins and triplets of the first generation are 

 from dihybrids and trihybrids of the second generation. 



All of the cultures were carried out in fairly rich garden soil and 

 some tillage was given, enough to keep down weeds and to stir the 

 soil several times, particularly during rosette development and during 

 the spring of the second season. Under these conditions the full ex- 

 pression and strength of characters is realized. On dry, moderately 

 poor garden soil plowed in the autumn, and seed sown broadcast in 

 October, and with no tillage, the annual plants were of smaller stature, 

 with few or no branches, and several characters which manifest them- 

 selves under more favorable soil conditions, or with tillage, are absent. 

 But even under the trying conditions stated, the majority of the stri- 

 king specific and contrast characters are present. The soil and tem- 

 perature conditions in a great part of North America appear to be 

 favorable to the growth of the evening primroses. In Central and 

 Western New York, at least, there appears to be no need of applications 

 of lime to the soil, which de Vries finds advantageous in Holland, since 

 the soil here is unusually rich in lime. 



For the technical diagnosis of the two parents employed in these 

 experimental studies, reference is made to the original description (see 

 Bartlett, 1913). A fair idea of the habits and characteristics of the 

 species can be obtained by a brief enumeration of the contrast "unit 



