340 



all obtained with certainty by determined experiments; the stock-form is always 

 found unchanged in the same culture with the variants. 



All the variants are from their origin as constant as their stock. 



The true factors which govern the variability in these experiments are still 

 unknown. 



2. By rapidly repeated re-inoculations and by other methods, normal form 

 and variants may be kept constant, as it seems for an unlimited length of time. 



3. All the variants vary in a way analogous to that of the normal form, 

 thus, the .7#/V7///s-variant produces an tf///tf/w-slimevariant, which must be consi- 

 dered as a gain-variant, and an albus-va.r\a.nt, which must be taken for a loss- 

 variant. 



The natural variety B. Kieliensis, which approaches the ffra/,w-variant, also 

 varies in an analogous way. The variation thus seems to be directed or ortho- 

 genetic. 



4. Gain-atavism in loss-variants and loss-atavism in gain-variants, can be 

 obtained whith certainty by dertermined experiments. Qualitative variants, too, 

 may give rise to atavism. 



5. The experimental variants of B. prodigiosus have not yet been found in 

 nature. From another bacterium, Bacillus herbicola, a variant, took rise which I 

 had before repeatedly isolated from nature and which I had taken for quite 

 another species. 



6. The variants of prodigiosus, and this holds good for many other microbes 

 also, differ from each other and from their stock forms in the same way as 

 closely related natural species or varieties do among each other. But their dis- 

 position to atavism is much more pronounced. 



7. The sub-variants, e. g. the rose variants of different colour-intensity, arise 

 in the same way as the chief variants and possess the same degree of constancy. 



