H. WORMALD 231 



Inoculation after 1 hour : one tube remained clear, the other became 

 turbid. 



Inoculation after 4 hours : one tube remained clear, the other became 

 turbid. 



Inoculation after 24 hours : both tubes remained clear. 

 >> jj 'J clays : ,, ,, ,, ,, 



111. Pleomorphism. 



In a former paper (46) the present writer stated that the celery-rot 

 organism when in vigorous growth measured 2-5;— 3-5 x 0-6 — 0-7 /u, with 

 double rods to 6'5/x in length; occasionally rods up to 16-5/a in length 

 were seen. A microscopic examination of the cultures treated with 

 the antiseptics as described in the preceding pages showed that in some 

 cases there was considerable divergence from the normal dimensions. 

 Observations showed that the behaviour of the organism in a culture 

 medium may be modified in one of two directions and varies with the 

 antiseptic with which the medium is treated. In the one case the 

 tendency is towards an inhibition of the process by which the individual 

 rod, on reaching a maximum length of about 6/x under favourable 

 conditions, divides transversely to form two rods which then separate. 

 The result of the lapse of this fission is that the rod increases in length 

 until it may be described as filamentous. In the other case fission 

 occurs while the rod is still short and the organism then resembles 

 a coccus. 



The celery-rot bacillus readily produces filaments in the presence 

 of carbolic acid, lysol, "lysol substitute," benzoic acid and salicylic 

 acid but only when the concentration is such that, while it does not 

 cause the further development of the organism to be suspended 

 altogether, it has the tendency not only to reduce the rate of growth 

 but also to behave as a factor limiting the amount of bacterial substance 

 ultimately formed. Thus filaments were generally to be found in 

 those tubes where the sediment was less bulky than in inoculated 

 control tubes. 



That the filamentous forms were derived from the bacillus intro- 

 duced at inoculation and were not the result of contamination was 

 clearly shown in the experiment in which tubes of bouillon were pre- 

 pared containing benzoic acid at three different concentrations, the 

 dilutions being obtained by mixing a saturated solution of the acid in 

 bouillon with sterile untreated bouillon ; the uninoculated tubes even 



