28 PROFESSOR LANKESTER. 



I assume it to be highly improbable that two or more speci- 

 fically distinct organisms -svould develope in a jar simulta- 

 neously, each tinted by the peculiar substance Bacterio-pur- 

 purin. At the same time the hypothesis that the various 

 forms observed do form a series belonging to a protean 

 species is confirmed by the observation of all conceivable 

 intermediate annectent forms. 



In my former paper I pointed out that the colour of the 

 films was liable to a certain variation, tending at one time 

 to blue and at another time to red. This variation in tint, 

 which I have now constantly observed, occurs in response 

 to definite variations in the conditions of growth, and is, I 

 believe, satisfactorily explained by the predominance of either 

 a blue or a red element of the complex colouring matter 

 Bacterio-purpurin. A certain brownish tinge is also observ- 

 able under some conditions, and is, I conceive, to be ex- 

 plained similarly. 



The bluer peach tint occurs when the growth is young, 

 and has been especially noticed by me when it occurs in 

 the form of a dense film (the plastids being arranged like a 

 mosaic or " tesselate," as in the mycoderma phase of 

 Bacterium termo, or in '^ catenular " series as seen in Plate 

 XXII, fig. 3, of my former paper), encrusting vegetable 

 matter such as dead twigs of trees. The redder colour is 

 assumed when the growth is more luxuriant, and is espe- 

 cially brilliant when the growth takes on the form of large 

 homogeneous discs (to be described in the present commu- 

 nication). In small specimens of these discs and also in 

 small biscuit-shaped plastids from Mr. Stewart's macerating 

 pan I have seen the colour of the greatest intensity, 

 only to be compared to the deepest magenta dye. Also 

 in the same conditions which furnish these most brilliantly 

 coloured examples are to be noticed others of a decidedly 

 brownish tint (Plate III, fig. 4, fig. 20). 



As will be mentioned below, the conditions under which 

 the homogeneous discs are developed have not been deter- 

 mined, but their occurrence seems to be due either to the 

 exhaustion of nutriment in the jar containing the growth, 

 or to the dying down of the green unicellular algae with 

 which Bacterium rubescens is usually most intimately asso- 

 ciated and upon which it may be dependent for oxygen, 

 and such services as the gonidia of the Lichen render to the 

 hyphal fungus. 



Variation in the size of the Plastids. — As was pointed out 

 in my previous paper, the form of the plastid of Bacterium 

 rubescens may vary greatly. It may be spherical, biscuit- 



