2 



mutants is in consequence not considered suf- 

 ficiently secure to be presented in detail at 

 present as proofs of mutation. The fact that 

 throughout these cultures the differentiating 

 characters behaved as units in the Mendelian 

 sense appears to me indubitable evidence that 

 the several elementary species have arisen 

 through mutation and hybridization. 



On the basis of allelomorphic differences 

 between different species these elementary 

 species of Bursa represent the closest possible 

 relationship between forms belonging to dis- 

 tinct types, since they are seen in most cases 

 to differ from each other by a single dis- 

 tinguishing unit or by two units at most. 

 The fact that Bursa Bursa-pastoris is every- 

 where recognized as a variable species, and the 

 presence of several distinct forms in each of 

 the localities from which material was derived 

 for these studies, show that these elementary 

 species generally grow in actual contact with 

 each other. There is neither geographical 

 isolation nor complete physiological isolation, 

 yet these nearly related elementary forme 

 maintain themselves absolutely distinct. This 

 should convince any one who may still enter- 

 tain any doubt regarding it that in the pres- 

 ence of Mendelian hybridization, no form of 

 physical isolation is necessary for the main- 

 tenance of closely related forms. Moreover, 

 these observations on Bursa show that Mende- 

 lian behavior is a strictly normal natural 

 process and in no manner dependent upon 

 the artificial conditions supplied by garden 

 practise. 



When a new form arises, differing from the 

 parent in one or several unit characters, these 

 new characters may be either dominant or 

 recessive to the corresponding character of the 

 parent. Less rarely they are neither domi- 

 nant nor recessive. The chances of survival 

 under these several possible conditions seem 



