242 FUNDAMENTALS OF CYTOLOGY 



lack of effective interbreeding, might represent early stages in the differ- 

 entiation of distinct species. 



The use of synaptic behavior in analyzing the relationships of poly- 

 ploid plants has already been described at page 219. 



Aberrations in Reproduction. — On several occasions cytology has 

 helped to determine the taxonomic status of certain plants by revealing 

 the presence of an atypical mode of reproduction. An instance of this 

 is seen in the Caninac section of the genus Rosa. Among these roses are 

 several which had been regarded as true species becaiLse of their constancy 

 of type and the lack of intermediate forms. It was found, however, that 

 they were polyploid in constitution and showed the most characteristic 

 type of hybrid chromosome behavior at meiosis. The cause of their 

 true breeding in spite of their meiotic irregularity w^as revealed in the 

 discovery that they are apomictic: their embryos do not arise from 

 sexual cells but from the nucellus by adventitious budding (page 146). 

 The prevalence of such a mode of reproduction in a genus thus tends to 

 preserve certain hybrids between the species and leads to the establish- 

 ment of a group of many constant and nearly similar units. Such 

 "agamic complexes" are known in numerous angiosperm genera, includ- 

 ing Ruhus, Citrus, Potentilla, Crepis, Taraxacum, Hieracium, Festuca, Poa, 

 and others. 



The Role of Chromosomal Changes in Speciation. — When it is 

 discovered that the number and the morphology of the chromosomes 

 can be used as characters in classification, one canriot pass directlj^ to 

 the conclusion that changes in these chromosomal features have by 

 themselves produced the differentiation of the taxonomic units in which 

 they are found. Visiljle chromosomal changes, like invisible gene 

 mutations, are factors in speciation, Ijut there are many conditions 

 that must be met if a newly formed chromosomal type is to become 

 established as a distinct species or subspecies in nature. The new type 

 must be physiologically suited to the habitat in which it arises. It 

 must be able to meet competition. It must either have sufficient sexual 

 fertility to maintain itself or be able to reproduce vegetatively. In 

 the latter case further modification through gene mutation would, of 

 course, be much slower than in a plant capable of sexual reproduction. 

 It would be stable and perhaps distinct in type, but not progressive. 



A sexually reproducing type should have some degree of isolation from 

 the parental type or types, for as long as it crosses freely with them the 

 production of intermediate types will prevent the attainment of specific 

 distinctness. The isolation permitting the new tj^pe to evolve inde- 

 pendentl}' through genie variation and selection may be of several kinds. 

 The type may be sexually isolated from the parental or other related 

 types by difference in flowering time, poor pollen tube growth after 



