287 



or whose colonies lack the so characteristic structure, and again others with that 

 structure, but wanting the denitrifying power. There are, too, intermediate forms 

 between the tough and the soft class, and I think it possible that they originate 

 from each other by mutation. 



That Bacterium sti/tzcri in the anorganic denitrifications possesses autotrophy, 

 follows from the above described experiment with the silicic plates. But this may 

 also be proved for colonies of organic origin, if only the right moment be 

 chosen for experimenting with them. In the organic plate cultures the autotrophy 

 of this species gets however rapidly lost. Only with quite fresh colonies, grown 

 on thiosulfate-agar plates, and transferred to the anorganic medium, just at the 

 time of their becoming visible a feeble but distinct anorganic denitrification could 

 be obtained, which continued during several days with the same degree of intensity, 

 only much feebler than the spontaneous denitrification. So it seems proved that 

 the autotrophy does not disappear as an indivisable factor, but may get lost in parts. 



That the autotrophy is really lost in the originally active colonies, is corro- 

 borated by the fact that not only the single colonies of the organic plates, but 

 likewise the combinations of the colonies of the different species are quite inactive. 

 Even all the colonies of broth-agar plates together, mixed with the undeveloped 

 germs lying between them, do not produce any denitrification in the anorganic 

 mixture. And this must be true for all the different species which produce anor- 

 ganic denitrifications and evidently possess the power of chemosynthesis in their 

 natural habitat. 



This form of variability is obviously analogous to that of the nitrate ferment, 

 which I formerly described 1 ) and as said called physiological species-formation. 

 In both cases a new elementary species is produced. It is remarkable that a 

 number of species or varieties living under the same conditions are subject to 

 this transformation, and that between the principal form and the one that has 

 completely lost its original character, some feebly denitrifying intermediate forms 

 are found, which may be compared to subspecies. 



Taking B. denitrificans as an example we can speak of B. denitrificans auto- 

 trophus and of B. denitrificans heterotrophus, the change being possible only in one 

 direction, at least with our present knowledge. 



This change is not a mutation in the accepted sense, as thereby the primitive 

 stock continues to exist wit the mutant under the same conditions under which 

 the latter was formed. Here on the contrary all germs change simultaneously, 

 so that in this case we have to do with a hereditarily constant modification, 

 comparable to the pleomorphy of many Fungi, and to a certain extent, to alter- 

 nation of generation. Comparable also to the production of somatic cells from 

 germ cells during the ontogony of higher animals and plants, a fact certainly 

 of general physiological signification. But modification and mutation are concep- 

 tions not sharply distinguishable and gradually related. 



') Ueber das Nitratferment und iiber physiologische Artbildung. Folia microbiologica, 

 3. Jahrg., Heft 2, pag. I, 1914. Recently I found that the ferment which produces nitrous 

 acid from ammonium salts behaves in the same manner and changes, when fed with 

 organic food into a saprophytuus non-nitrifying form. 



