ON A PEACH-COLOURED BACTERIUM. 37 



sively small " germs " from which Bacteria are supposed to 

 take their origin in experimental infusions. 



Reasons for assigning Bacterium rubescens to the Schizo- 

 mycetes and the Genus Bacterium. — It has been suggested to 

 me by some observers who have examined samples of the 

 growths which I have called Bacterium rubescens, that the 

 organism in question is not to be regarded as a colour-bearing 

 sjiecies of the group of Schizomycetes, but is rather to be 

 classed with some of the glocogenous forms of the so- 

 called unicellular Algse, such as Gloeocapsa. This view 

 appears to rest upon two facts — firstly, upon the assumed 

 absence of a motile phase of the plastids of our organism ; 

 secondly, upon the presence of a strongly developed colouring- 

 matter. 



The really important agreements between Bacterium 

 rubescens and the Schizomycetes appear to me to lie in the 

 following facts : that B. rubescens is one of the chlorophyll- 

 free Protophyta (consequently physiologically similar to the 

 Schizomycetes and Tungi), and that it exhibits the charac- 

 teristic vegetation forms of the Schizomycetes, namely, 

 spherical, biscuit-shaped, elongated or filamentous plastids, 

 devoid of nucleus, and aggregated either in linear series or 

 in sheet-like and massive fronds by means of a jelly-like 

 matrix. 



The assumed absence of a motile phase is an error. Pre- 

 cisely under the same conditions and with the same constancy 

 as is observed in the colourless Schizomycetes, active darting 

 movements of the plastids of Bacterium rubescens are 

 observed. It is a mistake to allow the commonly observed 

 active condition of the Schizomycetes, such as Bacterium 

 lineola and B. termo, to dominate altogether in our conception 

 of the essential features of the Schizomycetes. These species 

 exhibit, quite as commonly and abundantly as B. rubescens, 

 quiescent phases and gloeogenous aggregations distinguished 

 by Cohn as " zooglcea." 



There is very good ground for supposing that the common 

 micrococcus which appears before and with the biscuit-shaped 

 Bacterium termo in putrescent infusions, and which forms 

 *' rosary-chains " and other growths is a phase of the species, 

 whose active condition is distinguished as Bacterium termo ; 

 and we have similarly in B. rubescens — a micrococcus-phase 

 (the spherical plastids described in my previous communica- 

 tionj , and catenular rosary-chain aggregations belonging to a 

 species which in a certain phase of growth produces biscuit- 

 shaped plastids, such as those drawn in PL III, figs. 2 and 

 4, These are undeniably referable to the form-genus Bac- 



