418 



Cytological observations on Bact. coli 



In preparations made by the Bouin-Giemsa technique 

 but more heavily stained than for the demonstration of 

 cell boundaries, one-half of such a differentially affected 

 rod is seen to have partly disintegrated and stains faintly 

 pink, while the other half has the deep blue colour char- 

 acteristic of bacteria which were alive and growing at the 

 time of fixation. The half damaged rods show no sign of 

 a medial constriction but have the usual smooth contour 

 of resting bacteria; yet under the conditions of the 

 experiment each behaves as if consisting of two physio- 

 logically independent units. Presumably this indepen- 

 dence is not absolute and the fact that most of the affected 

 rods disintegrated as a whole implies some kind of 

 plasmatic connexion between the two cells of which 

 each is composed. 



(3) The relationship of the transverse septa to the ar- 

 rangement of the chromatinic bodies. As stated above, if 

 bacteria stained to show the cytoplasmic boundaries are 

 compared with bacteria of equal length stained for the 

 chromatinic bodies, it is found that the chromatinic 

 bodies lie between the boundaries. This alternation of the 

 two structures can also be demonstrated in the same 

 bacteria by a slight modification of the osmium-HCl- 

 Giemsa method. 



In ordinary hydrolysed preparations of B. mega- 

 therium and B. mycoides stained with Giemsa's solution, 

 red lines, corresponding in position with the transverse 

 partitions made visible by the sodium hydroxide treat- 

 ment are often visible in the differentiated, dehydrated 

 films mounted in Canada balsam (PI. 7, fig. 25), but such 

 lines are rare in similar preparations of Bact. coli and 

 B. mesentericus. If, however, films of these organisms 

 are mounted in water immediately after staining and 

 without differentiation the lines are clearly seen bisecting 

 the rod-shaped elements and intervening between re- 

 cently divided dumbbell bodies (PI. 6, figs. 15, 16, 17). 

 They are more distinct in B. mesentericus than in Bact. 

 coli in which they are usually too thin and faintly stained 

 for good photography (PI. 5, fig. 9). 



Normally cytoplasmic cell boundaries divide both 

 short rods and long winding filaments into a series of 

 cells, each with its own single or multiple, i.e. dividing, 

 chromatinic structure. 



If for some reason growth occurs without division of 

 the cytoplasm there is no orderly arrangement of the 

 chromatinic bodies. Long filaments are formed in which 

 no cell boundaries can be demonstrated and in which the 

 distribution of numerous chromatinic bodies is chaotic. 

 Such filaments (Text-fig. 3) arise spontaneously in cul- 

 tures of Bact. coli and Proteus, where they were also 

 observed by F. Neumann (1941), especially if derived 

 from old stock and they can be produced experimentally 

 (Robinow and Lea, unpublished observations) by certain 

 doses of ionizing radiations which do not interfere with 

 cytoplasmic growth (Lea et al. 1937). 



(4) The chromatinic structures of the cells of old cultures. 

 Even in old cultures kept from 18 hr. to 4 days at 37° C, 

 or for 2 days to several months at room temperature, 

 cells with well-developed chromatinic structures are never 

 completely absent, a fact already noted by Neumann 

 (1941). 



As stated above, such cultures contain three main 

 types of bacteria : small coccoid forms, short stout rods 



and small slender rods. The first predominate in old 

 cultures of Proteus vulgaris, while all three are repre- 

 sented in 18-24 hr. slant cultures of Bact. coli. When 

 stained for chromatinic matter by the OsO 4 -H01-Giemsa 

 method, old cultures present a variegated appearance. 

 A few colourless and structureless elements are obviously 

 dead remains (cf. cell above (e) in PI. 5, fig. 2), but most 

 bacteria of all three types contain one or more chroma- 

 tinic bodies (lower half of PI. 7, fig. 2). The chromatinic 

 structures are either deeply stained and centrally placed, 

 not touching the periphery, or less densely stained 

 dumbbell forms extending right across the cell (PI. 7, 

 fig. 22 b). The minute coccoid forms are best studied in 

 preparations mounted in water in which it is possible 

 to see the delicate layer of cytoplasm surrounding the 

 relatively large chromatinic structures; after dehydra- 

 tion and decolorization this marginal cytoplasm is often 

 invisible. 



Neumann (1941) has reported a diffuse Feulgen reac- 

 tion in all the cells of old cultures, I have found examples 

 of both diffuse and localized positive reactions in cells 

 from 18 to 24 hr. slant cultures, but these elements are 

 so small that interpretation is very difficult and I have 

 been unable to reach any definite conclusions about the 

 results of the Feulgen test at this stage. 



Preparations of old cultures made by the Bouin- 

 Giemsa method for the demonstration of cell boundaries 

 present a remarkable contrast to similar preparations of 

 young growing cultures. Instead of the bright, pure 

 blue coloration and conspicuous banded or chambered 

 appearance characteristic of the young bacteria (PI. 7, 

 fig. 24), the great majority of the old bacilli of all three 

 types stain a faint mauve or dull light blue and show 

 little or no sign of the differential staining which indicates 

 division processes (PI. 7, fig. 23). 



These differences in staining reaction reflect the 

 well-known change seen in living bacteria after sub- 

 culture, when the pale transparent cytoplasm of the old 

 cells acquires a brightly refractile appearance during the 

 so-called 'lag-phase' (which comprises all but the last 

 developmental stages shown in Text-fig. 1). 



Quantitative observations on five slant cultures of 

 Bact. coli incubated for 24-18 hr. at 37° C. were made 

 from films prepared by the Bouin-Giemsa method for 

 the demonstration of cell boundaries ; 500-600 bacteria 

 from each culture were examined. In these five cultures 

 active bacteria with deeply stained and clearly sub- 

 divided cytoplasm constituted only 4-7, 4-8, 5-5, 8-0 

 and 12-0% respectively of the total count. The re- 

 maining bacteria were faintly and uniformly stained 

 and the cell boundaries were either absent or indistinct. 

 On the other hand, osmium-HCl-Giemsa preparations 

 from two of the same cultures showed that 65-67 % of 

 the bacteria contained more than one chromatinic body. 

 Whether all the bacilli in which cell boundaries are indis- 

 tinguishable are really single cells, or whether the inter- 

 cellular boundaries are sometimes present but at this 

 stage are unstainable by the methods used, is not certain. 

 In spite of the diffuse Feulgen reaction and apparent 

 absence of cell boundaries in most bacteria in old 

 (18-24 hr.) cultures, the great majority of these or- 

 ganisms are fully viable. Neumann's view that there is 

 an ' early degeneration of the cells in well-grown cultures ' 

 is untenable when the changes induced in the chroma- 



77 



