406 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1914. 
pollination is a necessary condition of seed formation; Koelreuter 
demonstrated that characters from both parents appear in hybrid 
offspring; Amici, Pringsheim, Schmitz, and Strasburger showed how 
the mingling of parental qualities is made possible by the approxi- 
mation and mingling of parental protoplasms and nuclei. 
The sexuality which was first suspected and first experimentally 
proven in the seed plants has now been demonstrated in all groups 
of plants save the bacteria and their allies. The primary feature of 
the process, the union of two parental nuclei, is the same im all. 
The method of bringing together the two nuclei varies widely, this 
variation sometimes involving even the complete disappearance of 
externally recognizable sexual organs. During the evolution of 
plants old methods of accomplishmg the approximation of the nuclei 
have been discarded, and new methods have arisen. In the latter 
case a fusion of nuclei of closer kinship has often been substituted for 
the primitive one of more distantly related nuclei. This seems evi- 
dently the case, for example, in the apogamous Ascomycetes; perhaps 
also in the Basidiomycetes, and surely so in the cases of nuclear 
fusion in the prothallia and in the sporangia of apogamous ferns. 
In the process of fertilization, as we understand it at present, there 
are brought together two distinct sets of chromosomes, which in the 
nuclear divisions of the sporophyte, or 2X generation, are often found 
associated in pairs. The exact manner in which these chromosomes 
become paired, and the possibility of their attainmg any more inti- 
mate association, either in the resting reticulum or im synapsis, are 
not yet definitely determined. If, as is mdicated by Mendelian 
phenomena, and as demonstrated cytologically to the satisfaction of 
many workers, there is no loss of identity of the chromosomes in the 
sporophyte, then there is no very significant fusion at any pot in the 
life cycle in consequence of the sexual process. Members of the two 
sets of chromosomes may be interchanged or shuffled, probably at 
synapsis, and thus new sets or combinations be formed in the haploid 
nuclei at reduction. These new combinations, however, are still 
made up of the same discrete individual chromosomes. 
The essence of the sexual process then, as far as yet morphologically 
demonstrated, consists not of a real fusion, but merely of a temporary 
association, followed by a reassortment at sporogenesis of those 
ultimate, inheritance-bearing units—the chromosomes. 
