666 THE NODULE ORfiAXISM OF THE LEGUMIXOSiE, 



may be obtained in the usual way by inoculating a series of tubes 

 and pouring into Petri dishes. 



The colonies are circular, well raised from the surface and 

 white. The white colour may give place to a yellowish from 

 absorption of the colouring matter of the medium. In a pale 

 coloured medium the colonies are like drops of parattin or skimmed 

 milk; on the same plate both yellowish and white colonies have 

 been observed near one another. The yellowish was the older 

 colony, and apparently had absorbed all the free colouring matter 

 before the younger had made much progress. Although the 

 colonies do not litiuefy the gelatine, yet in some cultures a slight 

 liquefaction has been seen. This was obtained with a vigorous 

 culture growing upon a medium containing 6 per cent, gelatine 

 which, through prolonged heating during filtration, had lost some 

 of its gelatinising power. On the plates the colonies may consist 

 of many forms of the organism. Home colonies ma}- consist 

 entirel}' of short bipolar staining rods in the interior as well as 

 on the surface of the growth. Others again, even on the same 

 plate, may consist of these together with rods swollen at the ends 

 and exhibiting irregular staining, or with Y. satui'u-like, or 

 branching forms. 



The organism, generally speaking, is a capsulated l)acteriuni, 

 with rounded ends and stains irregulai'ly. The strong stains 

 such as fuchsin, unless the excess of colour is removed by alcohol, 

 show an irregular rod that may be more or less branched, while 

 the weaker stains as the blues show the protoplasm contracted in 

 places. The shorter bacterial forms are straight and stain at the 

 poles; the longer forms ma}^ be more or less bent, and show three, 

 four, five or more stained portions. The general shape varies 

 somewhat in the different media. In pe^Dtone-glucose fluid the 

 short bipolar staining rods predominate, while the substitution 

 of sucrose for glucose causes the irregular and l>ranching forms 

 to preponderate. On ordinary' meat-agar media the broken rods 

 appear to be thin in the middle; the addition of glycerine to the 

 meat agar causes some of the organisms to assume the long form, 

 the segregated protoplasm of which gives the rod a chlamydospore- 



