16 Mohr. 



Forked^ a secondary mutation. 



The fact that forked* is an allelomorph of orilinary forked attaches 

 a special interest to the manner of occurence of the mntation. The 

 presence of eosin and vermilion in the original modified - forked males 

 proves that the mutation which gave rise to the new character must 

 have occured in the eosin vermilion forked stock, i. e., in a fly which 

 beforehand was homozygous for ordinary forked. This means that the 

 new character is the result of a secondary mutation, being removed 

 from normal by two independent mutational steps. 



That the occurence of a mutation in a special locus does not prevent 

 the same locus from mutating once more, has been demonstrated in the 

 case of white at which locus mutations appear to take place more 

 often than at any other (see Morgan, Sturtevant, Müller and 

 Bridges '15). On several occasions the eosin allelomorph of white 

 has mutated to white and two of the latter cases were, as the one 

 here recorded, checked by control charactei-s. The secondary mutations 

 from eosin to white are progressive . i. e. they represent a further 

 departure from the wild tvi^e. They belong in other words to the same 

 t}ije as the one here recorded for the forked locus. The ultra -Bar 

 mutation recorded by Zeleny ('19) represents another case of the same 

 cathegory. 



The reverse step, a regressive mutation, i. e., a secondary mutation 

 in the direction back towards normal seems to be less frequent. It is 

 true that, as shown by May ('17), the dominant sex-linked character 

 Bar quite often has been found to revert to normal, but the exceptional 

 frequency of this particular change seems to indicate that special con- 

 ditions are probably acting. Eosin, the gene which has mutated to 

 white on several occasions, when first found by Dr. Morgan was the 

 result of a regTessive secondaiy mutation, since it arose in a white fly 

 (IVIorgan and Bridges '16). In the discussion of one of the gynandro- 

 morphs recorded by Morgan and Bridges ('19) the possible occurence 

 of a regressive mutation from white to normal is offered as an alter- 

 native explanation. Similarly the exceptional eye color found in a 

 gynandromorph (?) described in this paper (p. 21) may be explained as 

 the result of a regressive mutation of ruby^ to normal. But the latter 

 cases are also open to other explanations so that none of them are as 

 weU established as the ones in which the change went in the opposite 

 direction. 



