The Action of Gravity upon Bacterium Zopfii. 303 



photographs have been taken from plates so prepared. The organism 

 can always, however, be readily examined upon these gelatine slides 

 without fixing or staining. The method of cover-slip impressions we 

 have found not applicable in our case, as the growth is firmly 

 embedded in the gelatine ; it is useful, however, in the case of agar. 

 For reasons which will be given later, it is necessary to observe 

 under the microscope the ordinary test-tube growths in situ in the 

 gelatine. For this purpose we rapidly warm the test-tube to just pro- 

 duce liquefaction of the surface of the gelatine in contact with the 

 glass, and then slide it out upon a glass plate; a cover-slip is 

 placed over the surface of the growth, and it can then be examined. 

 Another method which we have employed, and one again which does 

 not lead to any disturbance of the growth, is the well known method 

 of drop cultures, but employing gelatine instead of a fluid medium. 



Fig. 1 is a photograph of a 36 hours' growth upon gelatine, the 

 result of a streak inoculation (see also fig. 12). The tube was kept 

 vertical. It is the characteristic growth, i.e., lines passing outwards 

 and upwards at an angle of about 45 : the appearance is typically 

 pinnate. In connexion with these vertical streak cultures we have 

 noticed this very curious phenomenon, that no matter how crooked our 

 original streak may be, yet the growth seems to form a new perfectly 

 median axis for itself, from which the rami start ; further, if instead 

 of a vertical streak we make three or four streaks upon the surface of 

 the gelatine at right angles to the long axis of the test-tube, the 

 growth still tends to form a median line. These facts, coupled with our 

 microscopic observations, pointed out to us that we might possibly be 

 dealing with a branching phase similar to that met with in Clado- 

 thrix dichotoma; and, indeed, if we refer to Zopf 's* drawings of Clado- 

 tlirix we do recognize a pinnate arrangement of the so-called false 

 branches of the filamentous organism. Billetf has termed this 

 branched state the "filamentous," meaning thereby the " etat d'ac- 

 croissement proprement dit de la plante ;" he has described it in 

 Cladothrix, Bacterium parasiticum, B. laminarice, B. urece, B. Balbiani, 

 and B. osteophilum. In Bacterium Zopfii we shall regard the branch- 

 ing growths as corresponding to the " filamentous phase." 



We soon found that we did not always obtain the characteristic 

 growth, even in the test-tube ; not only had the degree of sloping, as 

 mentioned in the commencement, a considerable effect upon its 

 regularity, but the temperature also exercises a marked effect. We 

 obtained the best growths in a temperature of about 21 C. kept 

 constant. In this way a perfect growth might be obtained in 

 24 hours, whilst if that portion of the test-tube which corresponded 



* Schenk's ' Handbuch der Botanik.' 



f ' Contrib. a I'fitucle de la Morphologie et du Developpcment des Bacteriacecs.' 

 Paris, 1890. 



