CYTOLOGY AND TAXONOMY 241 



(lept'iicLs piimaiily upon the homology, or genie similarity, of the chromo- 

 somes concerned, any lack of synapsis therefore indicating a hick of close* 

 relationship between them (except when it is induced b.y environmental 

 causes or special mutant genes). In diploid hybrids, synapsis ranges all 

 the wa}'' from perfect success to total failure, and in general, though not 

 always, fertility varies with the degree of success. Reverting to Zea 

 and related genera mentioned above, it is found that when maize is 

 crossed with either annual Mexican teosinte or teosinte from northern 

 Guatemala the hybrid shows normal synapsis (Fig. 162). In hybrids 

 between maize and southern Guatemalan teosinte, synaptic association 

 shows a few abnormalities due to differences in the linear arrangement of 

 homologous elements. In maize-gamagrass hybrids synapsis occurs 

 only in a few regions of the chromosomes, indicating a low degree of 

 homology-. Similar beha^'ior is observed in teosinte-gamagrass hybrids. 

 In triple hybrids containing genomes of all three genera, the Zea and 

 Euchlaena chromosomes pair, leaving the Tripsacum chromosomes 

 unpaired except for an occasional trivalent. These cytological findings 

 are a valuable supplement to taxonomic and breeding evidence for 

 generic relationship. They afford a visible measure of the homology 

 and arrangement of the genes, especially since these plants show the 

 process so clearly. 



Evidence comparable to the above is available in certain animal 

 groups also. It has been possible to make the most minute comparisons 

 of chromosomes of related species in the Diptera because of the giant 

 chromosomes in their salivary glands. Years ago the ordinary somatic 

 chromosomes of various species of Drosophila were compared, but much 

 more can now be learned from a comparison of their salivary-chromosome 

 maps. In intraspecific hybrids the intimate association of parental 

 elements in the salivary-gland chromosomes permits an even more 

 precise comparison. In this way a point of special interest has been 

 brought out with respect to certain geographical strains of Drosophila 

 pseudoobscura. By studying the chromosomes of hybrids between these 

 strains it was found that three of them have the genetical elements in 

 chromosome III arranged in three different ways as the result of inver- 

 sions: (1) ABCDEFGHI, (2) AFEDCBGHT, (3) AFEHGBCDI. In 

 these arrangements there is a clue to the historical sequence of the races. 

 Two successive inversions could easily give the sequence (1) -^ (2) -^ (3) 

 or the sequence (3) — ^ (2) — > (1). Also, (2) might have been the pre- 

 cursor of both (1) and (3). On the other hand, it is very highly improb- 

 able that an alteration directly from (1) to (3) or vice versa would occur 

 at one step through some more comj^lex aberration. Since the "struc- 

 tural hybrids" between such inversion strains in Drosophila tend to be 

 sterile, it has been suggested that such strains, because of the consequent 



