182 BACTERIA IN THE SPUTA 



With the above-mentioned stumps are intermixed almost all 

 the forms of bacteria, bacilli, spirilla, etc., besides a large number 

 of very varied small chains, after the types <:, d^ k, i, x, and x 

 (Fig. 2). 



The larger size and the facility to colour of these masses and 

 stumps, with regard to fertile filaments, would depend, I think, 

 upon a very natural cause. The friction, removing the points of 

 the filaments and the fertile filaments, seems to impart a greater 

 development to the remaining stems in the sense of thickness ; 

 as, for instance, it happens in the pruning of trees, through the 

 retrocession of the ascending sap. 



In the present case, the retrocession of the germinal matter 

 would be the cause of the consecutive increase of the sheath and 

 of the gemmules contained in it. Thus might be easily explained 

 the appearances, so various, of the stumps and their fragments, 

 compared with the fertile filaments ; appearances which induced 

 Miller to make of residual filaments two distinct species, under 

 the name of Leptothrix buccalis maxima and Bacillus buccalis 

 maximus, in opposition to our fertile filaments, which for Miller 

 constitutes a third species, the Leptothrix inno??ii?iata. 



It remains, then, to be seen whether the longer filaments, sunk 

 down and intertwined, of Bacillus buccalis maxifnus (Fig. 13 of 

 Miller) may not concur, with the above described chains, to form 

 that kind of mycelium or creeping vegetation, upon which is based 

 the aerial vegetation of this microphite. 



From the tiny islands of the stumps in question, prepared in 

 the manner already described in Section 2, spring at last the fruc- 

 tifications. The bigger filaments, that we would call woody, gra- 

 dually become thinner and pale, showing in their interior countless 

 granules or parietal gemmules. These are the fertile filaments. 

 They may spring either from the proper filaments, with continuous 

 contour, or from the little chains after type d (Fig. 9), which are 

 seen occasionally on the top of those filaments ; and in this last 

 case, instead of a gradual thinning, may abruptly pass from the 

 chain to the fertile filament, as shown in the figure. 



It is by no means uncommon to find parietal gemmules, as 

 Robin mentions the circumstance, and gives a first drawing in 

 Fig. 2h of his volume of 1853, magnified to 800 diameters.* 



* Robin, Hist. Nat, des Vegetaux, etc., 1853. (See Bibliography.) 



