DEPARTMENT OF EXPERIMENTAL EVOLUTION. 93 



Prolific Black Wax have allelomorphs for a dark yellowish-brown color, which 

 can be demonstrated by appropriate breeding-tests, and Ne Plus Ultra, which 

 is dark orange, carries light yellow as a latent character. The crosses of the 

 latter variety lend support to the "presence and absence" hypothesis. In dis- 

 cussing these results, four different types of latency have been recognized, 

 namely, (a) Latency due to separation, as when two allelomorphs which are 

 jointly necessary to the production of a visible character do not occur to- 

 gether in the same soma. This condition is very frequent in the case of albinos, 

 (b) Latency due to combination, in which the cooperation of two allelomorphs 

 in some way destroys or prevents their characteristic manifestations, thus ren- 

 dering the individual possessing both of the allelomorphs in question quite 

 indistinguishable from individuals which possess neither of them, (c) Lat- 

 ency due to hypostasis, the very common case of complete invisibility of one 

 characteristic in the presence of another, the latter characteristic not being 

 perceptibly modified by the presence of the former, (d) Latency due to fluc- 

 tuation, when characters which are normally present are temporarily sup- 

 pressed through malnutrition or other similar causes. The first three of these 

 types of latency is exemplified by the hybrid beans, and each results in the 

 production of definite ratios which differ from the typical Mendelian ratios 

 without in the least affecting the fundamental principle of segregation of 

 "pure" gametes and their union according to the laws of chance. 



Studies on the effects of cross- and self-fertilization in maize have given 

 further evidence that an ordinary field of Indian corn consists of a series of 

 more or less complex hybrids among numerous elementary species or bio- 

 types, and that the apparently injurious eft'ects of self-fertilization are due 

 simply to the unfavorable comparison of pure strains with their hybrids, and 

 of less complex hybrids with more complex ones. Self-fertilization sooner or 

 later reduces any pedigree to the condition of a pure strain by eliminating its 

 hybrid elements. According to this hypothesis, when the strain is once re- 

 duced to a pure state, no further deterioration should result from continued 

 self-fertilization. All the evidences available at this time appear to support 

 this proposition. 



In the common large-flowered garden sunflower, or Russian sunflower, there 

 are two elementary forms, one of which is practically unbranched, the other 

 having several to many lateral branches, particularly on the upper portion of 

 the central axis. When the unbranched type possesses a few branches, as it 

 does occasionally, these are near the base of the central axis. In crosses be- 

 tween these two forms, the branched type is dominant over the unbranched, 

 the heterozygotes being quite indistinguishable from their branched parent. 

 Segregation is perfect, regardless of the fact that the branching habit is rather 

 easily modified by environment. In crosses between the Russian sunflower 

 and the semi-wild Helianthus amiiius of the prairies, the behavior with respect 

 to branching is quite anomalous, and it is expected that this may throw some 

 light upon the phylogenetic relations of these two forms. 



