416 



Cytological observations on Bact. coli 



as stated above, it may already occur after the first 

 division of the chromatinic body. 



When plump bacillary forms (type 2) are reactivated 

 by transference to a fresh nutrient medium the two, 

 three or four chromatinic structures which each contains 

 first stretch and enlarge, and then divide into dumbbell 

 bodies which become symmetrically distributed along 

 the bacillus. In this way the same type of bacterium 

 with two or four resting or dividing chromatinic bodies 

 is produced, as that derived from the coccoid forms. » 



When the slender rod forms {type 3) are about to 

 resume their growth activities, their chromatinic struc- 

 tures are at first arranged in a long, almost homogeneous 

 core (PI. 5, figs. 1, 2, and fig. 9, top corners). During the 

 first 2 hr. after subculture, the chromatinic core, which 

 is presumably a column of tightly packed chromatinic 

 elements, breaks up and forms groups of regularly spaced 

 dumbbell bodies between which cytoplasmic boundaries 

 then develop. Eventually large rod forms result which 

 have the same structure and nuclear complement and 

 follow the same development as the rods directly derived 

 from the coccoid elements. For example, in PI. 5, fig. 1, 

 the very large bacterium, above (a), has probably de- 

 veloped from a rod with three chromatin masses like 

 that at the extreme right edge of the same figure ; this 

 would explain the unusual asymmetrical distribution of 

 the chromatinic bodies in the large bacterium. 



The redistribution of the chromatinic matter and its 

 differentiation into dumbbell bodies in the slender rod 

 forms of reactivated cultures of Bad. coli and Proteus 

 vulgaris is essentially the same process as that previously 

 described in the vegetative cells from old cultures of 

 B. mycoides after transference to fresh nutrient medium 

 (Rob'inow, 1942). 



During the second hour of incubation the bacteria, 

 which are now predominantly rod-shaped (PI. 6, figs. 

 20, 21) begin to react differently to Giemsa's stain. The 

 chromatinic bodies in the early stages are stout and very 

 deeply stained (cf. PI. 5, figs. 1, 2 and 7) and the cyto- 

 plasm assumes a reddish colour; in the older bacilli the 

 dumbbell bodies become slender and in between divisions 

 stain comparatively faintly, while the cytoplasm, though 

 it may show a reddish tinge while still in the staining 

 solution, becomes light blue in the acetone-xylol mixtures 

 by which it is now very easily decolorized.. 



At all stages of growth division of the chromatinic 

 bodies may result either in two single dumbbells which 

 separate completely before beginning a fresh process of 

 division (PI. 5, figs. 5-7), or the daughter bodies may 

 start to divide before they have yet completely sepa- 

 rated, as described above for the first division of the 

 chromatinic structure. From the second hour onwards, 

 the greater transparency of the cytoplasm renders the 

 differences between faintly stained, slender, resting 

 dumbbells and deeply stained, stout, dividing dumb- 

 bells much more distinct than in the earliest growth 

 stages when the cytoplasm is more heavily stained. 



When two plump dumbbell bodies of symmetrical 

 build are closely contiguous it is impossible to see whether 

 their long axes are parallel or at right angles to the long 

 axis of the cell. (For an illustration of this difficulty, as 

 well as its solution, in the case of a large spore-forming 

 bacterium see PI. 6, figs. 15 and 16.) The latter arrange- 

 ment is the usual one; Neumann (1941) describes cases, 



however, in which one or two dumbbell bodies lie near 

 the surface of the bacterium and parallel with the long 

 axis of the cell. As a rare occurrence I have confirmed 

 this observation in my own cultures. 



As the culture gets older the bacteria become shorter 

 and narrower and their chromatinic structures increas- 

 ingly difficult to resolve (PI. 6, figs. 20, 21); at the edge 

 of isolated colonies, however, bacteria with well-differ- 

 entiated chromatinic structures can be demonstrated 

 until the growth of the colony ceases. 



At the edge of colonies of Proteus vulgaris growing on 

 a suitably moist solid medium, periods of intense repro- 

 ductive activity alternate with the well-known pheno- 

 menon of swarming. The chromatinic structures in the 

 long filaments of the swarming zone are very regularly 

 spaced (PI. 6, fig. 14) but the individual elements are 

 less clearly resolved than in the short rods (PI. 6, fig. 13) 

 which predominate in the marginal zone of a Proteus 

 colony between the periods of swarming. The long fila- 

 ments may divide by breaking into pieces of equal length 

 (PI. 6, fig. 14 b), but usually fragments of variable length 

 break off from one end. The cause of the periodic changes 

 in the periphery of Proteus plate cultures remains obscure 

 (cf. Russ-Muenzer, 1935). 



(2) The composite structure of rod-shaped bacteria 



The mono- and multinucleate elements described 

 above are really single and multiple forms of a basic 

 building unit possessing a single chromatinic body (PI. 5, 

 fig. 10a). The simplest form of this unit is represented 

 by the coccoid elements found in old cultures and in the 

 marginal zone of Proteus colonies between the periods 

 of swarming and the composite form by the rod-shaped 

 bacteria which predominate in young cultures. Although 

 the rod-shaped bacteria appear homogeneous in ordinary 

 preparations, their composite structure can be clearly 

 demonstrated in the following ways: 



(a) Staining (Bouin fixation). When bacteria are fixed 

 through the agar with Bouin's fluid (see p. 414) and 

 stained for not more than a few minutes with Giemsa 

 solution, the nuclear bodies and the cell wall are not re- 

 vealed but most of the rod-shaped elements show a regular 

 pattern of dark transverse lines and their tips also are 

 deeply stained (PI. 8, figs. 28, 29; PI. 7, fig. 24). The 

 lines are absent in the smallest elements though the 

 poles of the cell may be deeply stained (PI. 8, fig. 28 a). 



The relative position of the dark lines and the chroma- 

 tinic bodies can be determined by comparing bacteria 

 treated by the Bouin-Giemsa method with elements of 

 equal length in preparations made by the Os0 4 -HCl- 

 Giemsa technique (cf. PI. 6, figs. 17, 18). This comparison 

 shows that the dark lines always occupy the space 

 between the chromatinic structures, which explains why 

 the smallest elements with only one chromatinic body 

 lack transverse lines. 



Young bacilli in Bouin-Giemsa preparations have the 

 following appearance (PI. 8, figs. 28, 29). Usually a very 

 broad and dark double band runs across the middle of 

 the rod with a fine line symmetrically placed on either 

 side of it so that the rod is subdivided into four equal 

 sections. In long bacteria there may be several thick 

 double lines evenly distributed on either side of the 

 broad central band, with a fine line midway between 

 each pair of thicker ones. 



175 



