498 Conclusion. 



the garden and allowed them to bear fruit in a state of 

 perfect isolation. We can be guided, to a certain extent, 

 in our choice of species by observations in the field ; either 

 by finding new varieties or subspecies; or by a certain 

 richness in partial variations, or by so-called monstrosi- 

 ties. The latter are due to latent potentialities which are 

 manifested from time to time on isolated twigs, leaves, 

 etc. It is natural to conclude that where such latent 

 potentialities occur in unusually large numbers, others 

 may be expected and, amongst them, those that we are 

 looking for. 



I have conducted a considerable number of such ex- 

 periments, both before and after I began my main ex- 

 periment, particularly with species of our indigenous 

 flora, on such a scale as the amount of seed harvested 

 would allow. For example, I sowed seeds of Capsella 

 Bursa Pastoris, Sisynibriuin Alliaria, Daunts Carota, 

 Cvnoglossuni officinale, Verhasciun thapsifornie, Aster 

 Tripolium, Bid ens cernua, Thrincia hirta, Crcpis biennis, 

 Ccntaurca nigra and a whole series of other wild species. 

 They were mostly forms which attracted my attention 

 by the possession of fasciations, concrescences, or other 

 kinds of abnormality. I cultivated the monstrosities for 

 longer or shorter periods of years in order to test their 

 hereditary nature. 



Almost all the species proved themselves to be im- 

 mutable.^ I conclude from this that most of the wild 

 species in our neighborhood happen to be in an immutable 

 state. In other places the same species may of course 

 be mutable,^ for, according to the theory, mutability is 



^ See also the first part of the second volume. 



^This seems to be true of Capsella Bursa Pastoris, near Landau; 

 see SoLMS Laubach^ Bot. Zeitung, 1900, October part. 



