MENDELISM AND MUTATION 349 



and T. compactum) 12 (hexaploid). He concludes thai the one-grained 

 wheats are the ancestral forms from which the emmer and spell wheats 

 have arisen through changes in chromosome number. This is precisely 

 the conclusion which Schulz (1913) and Zade (1914, 1918) had reached 

 on other grounds. Kihara (1919) found further thai by crossing emmer 

 and spelt wheats fertile hybrids with 35 chromosomes could be obtained, 

 and that these in future generations produced forms wit h varying chromo- 

 some numbers because of the irregular manner in which the chromo- 

 somes are distributed at the time of reduction. (See p. 253.) &£ a 

 general rule hybrids produced by crossing forms with different chromo- 

 some numbers are sterile, but when they are fertile and their chromosome 

 number is odd there is usually an irregularity in genetic behavior for 

 several generations until the number again becomes settled. In some 

 cases the number thus settled upon is that of the original ancestor with 

 the lower number (the (Enothera mutants of deVries and Stomps cited 

 above), whereas in other cases (Triticum) it is that of the ancestor with 

 the higher number. The manner in which many such changes in number 

 occur is not yet known. 



A study of the chromosomes in the genus Crepis has been made from 

 this point of view by Rosenberg (1918, 1920). He finds four species, 

 including C. virens, with three pairs of chromosomes, eight species with 

 four, four species with five, one species with eight, one species with nine, 

 and three species with 21. In Crepis the chromosomes differ markedly 

 in size, and Rosenberg concludes that the species with three, four, and five 

 pairs have arisen through such irregular distribution of the smaller chro- 

 mosomes as has actually been observed in the maturation division-. 

 together with recombinations occurring at fertilization. The segmenta- 

 tion of the larger chromosomes of the complement is not thought to 



occur. 



In an extensive investigation of the chromosomes of Zea Mays 

 Kuwada (1919) has found cytological confirmation of the conclusion of 

 Collins (1912) that -this species, which for some years has played a 

 conspicuous role in genetical investigations, is in all probability a hybrid 

 between Euchlcena ?nexicana (teosinte) and some other unknown form 

 belonging to the nearly related Andropogonese. Owing to their in- 

 equality in size Kuwada is able to distinguish what he considers to be 

 the chromosomes of the two supposed ancestral derivations in the cells 

 of certain races of maize. Thus gemini with components of unequal 

 size are frequently observed in the microsporocytes. 



Animals. — The most complete descript ion of the chromosomes in alarge 

 numberof closely related animal species is thai given by Metzi 191 1. 19166) 

 for the Drosophilidse. In about 30 species Met z has identified no less than 

 12 main types of chromosome groups, all but one of them being found in 

 the genus Drosophila. In Fig. 137 are shown diagrannnat ically the 12 



