148 CLELAND 



being revealed, and this in a genus which has for many years been the despair 

 of taxonomists who have tried to ferret out the relationships by strictly taxo- 

 nomic methods. 



In plants another cytological phenomenon has proved to be of great im- 

 portance in determining relationships. It not infrequently occurs in plants 

 that the chromosome number becomes altered. Chromosomes may occasionally 

 be lost, or additional chromosomes may be added. Of greatest significance are 

 the cases where the ordinary or diploid chromosome number becomes doubled, 

 producing what is known as a polyploid. It has been found that polyploidy 

 has been a major factor in plant evolution. A polyploid does not produce 

 fertile progeny easily when crossed with its parental diploid, because the 

 progeny will have an odd number of chromosome sets and will fail to a large 

 extent to produce sperms or eggs with complete sets of chromosomes. A poly- 

 ploid, therefore, has a degree of reproductive isolation which enables it to 

 carry on an independent existence and to evolve along its own line. A very 

 large number of species of plants show evidence of chromosome doubling 

 in their evolution — in fact, doubling has occurred more than once in the 

 ancestry of many plants. A curious fact has been discovered in the course of 

 studies of polyploidy: if doubling occurs in a plant whose parents were very 

 closely related, the resultant polyploid is likely to be more sterile and less 

 able to maintain itself than if its parents had been very unrelated. This fact, 

 the reason for which we cannot go into here, is of great practical value. It 

 is often advantageous to combine the desirable traits of different varieties or 

 species, but when one crosses these, the resultant hybrid proves to be sterile — 

 the chromosomes of the two species are too unlike to pair and separate 

 properly during the process of germ-cell formation. Such a cross would not 

 be able to propagate itself, therefore, were it not for the possibility of poly- 

 ploidy. If one induces the chromosomes to double in such a hybrid (and 

 this can be done easily with the use of colchicine or other chemical), the 

 resultant polyploid, which will still possess the desired combination of char- 

 acters, will be found to be perfectly fertile and capable of passing its desirable 

 combination of genes to succeeding generations. The wider the cross and the 

 more sterile the hybrid, the more likelihood that a polyploid derived from it 

 by chromosome doubling will be perfectly fertile, and vice versa. 



In the course of evolution, chromosomes are apt to suffer various major 

 or minor alterations in structure. They may lose small segments, or certain 

 segments may become duplicated. They may experience inversions, or they 

 may exchange segments with other chromosomes. As a result, chromosomes 

 in races which were originally derived from a common ancestor may become 

 very different in structure, so different that if they are brought into the same 

 plant by appropriate crossing they can no longer associate in pairs as corre- 

 sponding chromosomes are supposed to do at the time when germ cells are 

 produced. Cytologists can study the behavior of chromosomes in hybrids and 



