THE NUCLEAR APPARATUS 21 



in certain bacilli, and a permanent axial spiral in others, both of which he regarded as 

 being nuclear in character. In certain large micrococci he observed well-differentiated 

 and homogeneous bodies which took up nuclear stains, divided before cell division occurred, 

 and had the essential characters of nuclei. Reviewing his own results and those of other 

 investigators, DobeU concluded that bacteria are nucleated cells, in which the nuclear 

 apparatus is highly variable in form, appearing sometimes as scattered granules, some- 

 times as a spiral filament, and occasionally as a differentiated nucleus similar to that in 

 cells of higher organisms. Lewis (1932, 1934, 1940) doubts the reality of nuclear spiral 

 filaments, believing rather that they represent fine strands of cytoplasm compressed by 

 an abundance of volutin granules or fat globules. 



In recent years, the evidence for the existence of a discrete nuclear apparatus in bacteria 

 has been accumulating. HoUande (1934) (see also Hollande and Hollande 1932) differ- 

 entiated three structures in a number of bacilli, which he terms nucleosome, paranucleo- 

 some and metanucleosome. The paranucleosome is eosinophilic, and the metanucleosome, 

 which surrounds it, is basophUio. Both are closely associated with the nucleosome, 

 which is a minute spherical granule staining blue with eosinate of methylene blue. The 

 nucleosome divides by elongation, constriction of the central portion into a thread, and 

 finally separation into two daughter nucleosomes. Later workers confirm Hollande's 

 work to the extent of recognizing bodies, usually referred to as nucleoids, that take part 

 in cell division. Dombrowsky (1936) described two concentrations of protojjlasm in 

 Bad. coli, which appeared prior to division ; the line of division occurred between them. 

 In young cells of another coliform bacillus Rosea (1937) found a chromatinic substance 

 developing into a distinct " nucleus " and noted the appearances resembling mitotic 

 figures. Stille (1937) was able to demonstrate Feulgen-reacting bodies in a large number 

 of bacterial species after they had undergone gentle acid hydrolysis. Young vegetative 

 forms contained two bodies, though cells containing four bodies a2)peared prior to division. 

 Similar mUd hydrolysis reveals nucleoids in Bad. coli and Salm. parati/phi B (Piekarski 

 1937) and in Proteus vulgaris (Neumann 1941). Piekarski distinguished a primary foi-m 

 with two bodies, and a secondary form, usually found in older cultures, with one only. 

 The bodies were Feulgen positive, and absorbed ultra-violet light of a wave-length known 

 to be strongly adsorbed by thymonucleic acid (Piekarski 1938). Later it was found that 

 all the bacteria studied contained at least one nucleoid (Piekarski 1939, 1940, Piekarski 

 and Ruska 1939 a, b). In some cases the nucleoids were demonstrable in electron micro- 

 graphs. It may be noted here that the electron microscope has so far not revealed very 

 much of the nuclear structure of bacteria (see, for example, von Borries et al. 1938, Lembke 

 and Ruska 1940, Mudd et al. 1942, Knaysi and Mudd 1943). If any nuclear material 

 is present, it must either be masked by other matter more opaque to the electron beam, 

 or have the same electron scattering jiower as other cellular constituents. 



Robinow (1942, 1944) has extended these studies of the nuclear apparatus revealed 

 by preHminary acid treatment of the bacteria. Robinow's results not only confirm and 

 extend those of Stille and Piekarski, but clarify the relation of the Feulgen-positive 

 (chromatinic) bodies to the growth of the bacteria. In old bacteria, the chromatinic 

 bodies are small, and difficult to demonstrate ; when the bacteria are transferred to a 

 fresh nutrient medium, the bodies increase in size and stain more deeply, usually having 

 the form of short, dumbbell shaped rods. Longitudinal divisions of these rods precedes 

 the division of the cell, though several divisions may take place before the evidence of 

 cellular division is evident by ordinary methods of examination. Growing baciUi may 

 therefore contain two or more pairs of chromatinic bodies (Fig. 5). By special staining 

 methods these multinucleate bacilli appear to be banded, indicating the existence of 

 multiple cell units within the single bacillus. In the case of B. megatherium the composite 

 nature of the cell can also be demonstrated by a plasmolysing treatment (see p. 28 below). 

 The possibility that single bacUlary rods may in fact be multicellular must be borne in 

 mind in the interpretation not only of morphological studies on bacteria, but also of the 

 phenomena of heredity that they display. 



