A RECESSIVE CHARACTER AND SELECTION 175 



some of the variability was due to modifying factors, but later 

 Castle ('16) claimed to have disproved that modifying factors 

 were concerned because there was no decrease in the rapidity of 

 change after seventeen successive selections. For this and other 

 reasons Castle (Castle and Wright, '16; Castle, '17) rejects the 

 hypothesis of modifying factors. 



On the other hand, Marshall and Muller ('17) found that the 

 factor for balloon wing which is variable showed no evidence of 

 contamination after having been in association with its allelo- 

 morph for more than fifty generations. Morgan ('15) concluded 

 from the evidence furnished by the character, abnormal abdo- 

 men, that no contamination of genes or gametes occurred. The 

 expression of this character is dependent upon the environment, 

 rendering it possible for the individual to be normal in appear- 

 ance when the character is either heterozygous or homozygous 

 for abnormal. If the environment is favorable, a heterozygous 

 individual will be abnormal. In none of his work did Morgan 

 find any evidence to support the theory of gametic contamination. 



Either gametic contamination or modifying factors would cause 

 characters to vary, and in both cases progress by selection would 

 be possible. Zeleny and Mattoon ('15) found that the number 

 of facets in 'bar eye' Drosophila could be increased or decreased 

 by selection. They are inclined to the belief that the variabily 

 in the number of facets is due to 'differences in factorial com- 

 position.' MacDowell ('15) showed that selection was effective 

 in increasing the number of extra bristles in Drosophila. He 

 says: "The hypothesis of accessory factors wiU explain aU the 

 facts, and that of modification of a Mendelian factor may be 

 employed to interpret most of them." 



Morgan ('16) reports that in his laboratory Muller selected for 

 shorter wings in the truncate stock of Drosophila and obtained 

 individuals at the end of the experiment with much shorter wings 

 than those with which he started. He showed by means of link- 

 age experiments ''that at least three factors were present that 

 modified the wings." Dexter ('14) found that the variability 

 of beaded wmg was affected by a modifying factor in the second 

 chromosome, which could be eliminated or preserved, thus making 



