82 



siasm of the investigator to the highest pitch and induced him to 

 plan a series of observations lasting through the following four- 

 teen years, which entailed an enormous amount of tedious me- 

 chanical work. 



The experimental investigations which followed were organized 

 with* an attention to detail that left nothing to chance and re- 

 duced the sources of error to a minimum. Without recounting 

 all of the various features of the technique employed it will suffice 

 to say that most rigid methods of isolation of specimens and 

 regulation of pollination were used, a fact that adds much value 

 to the results which were achieved. 



The culture of the main species was carried through succes- 

 sive seasons from absolutely pure crops of seeds, and the deriva- 

 tive forms were cared for in the same manner. Onagra biennis 

 was found to be in a state of mutation and to reproduce every 

 year of the experimental tests a number of forms which might 

 be designated as constituting new species. Nine species were 

 thus found to spring from 0. biennis, being designated as gigas, 

 albida, oblonga, rubrinervis , ^nanella, lata, scintillans, clliptica, Icp- 

 tocarpa, brevistylis and lacvifolia. Mutants or derived species also 

 were found to be in a state of mutation to some extent, and lacvi- 

 folia gave rise to spatludata in addition to some of the forms 

 derived from the parent type. 0. lata in a similar manner gave 

 rise to sublinearis and subovata in addition to some of the forms 

 also derived from the parent type. The following table will 



A Group of Derivatives from O. biennis 



Name of Species 

 oblonga. rubrinervis. biennis. nanella. lata, scintillans. 

 O 



Generation. 

 VIII. 1899 



gigai 



■. albida. 



5 — 



obit 

 I 



VII. 



1898 







9 



VI. 



1897 





11 — 



29 



V. 



1896 





25 — 



135 



IV. 



1895 



I — 



15 — 



176 



III. 



1890 



-91 







II. 



1888- 



-89 







I. 



1886- 



-87 







— 8 — 



1700 - 



1 



- 21 





1 

 3000 - 



- 11 





1800 - 



1 



- 9 - 



- 5 



1 

 8000 - 



- 49 - 



- 142 



14000 - 



- 60 - 



- 73 



I 0000 - 



1 



3 ~ 



- 3 



1 

 15000 - 



1 



- 5 - 



- 5 



1 

 9 







