414 



Cytologicdl observations on Bact. coli 



by Giemsa (1912), is not indispensable, but the contrast 

 between chromatin and cytoplasm in the stained pre- 

 paration is much sharper and the staining is more uniform 

 throughout the preparation than after osmic vapour 

 fixation alone. 



For the demonstration of cytoplasmic cell boundaries 

 (vide infra) E. Klieneberger's (1934, 1942) method of 

 fixation through the agar with Bouin's fluid was used in 

 a simplified form. The cover-slips bearing the agar strips 

 were left for 30-45 min. or longer in a large volume of 

 the fixative after which the agar was removed and the 

 preparation stored in 70 % alcohol. 



(2) Proteus and Bact. coli 



After many disappointments the making of wet smears 

 was completely abandoned in favour of impression pre- 

 parations or of dry smears from material fixed in osmic 

 vapour while still on the agar. To make an impression 

 preparation, a square was cut from the agar plate, not 

 larger than could easily be handled by the tip of an 

 iridectomy knife and accommodated on a £ in. sq. 

 cover-slip. This square was exposed to osmic vapour for 

 2-3 min. and was then placed face downwards on a 

 clean, dry cover-slip. After removal of the agar, the 

 film of fixed bacteria deposited on the cover-slip was 

 dried, treated with warm sublimate alcohol as described 

 above (§ A (1)), and finally rinsed and stored in 70% 

 alcohol. Klieneberger's method of fixation through the 

 agar with Bouin's fluid was also used (see § A (1)). 



Not only are the bacteria in the impression films 

 broader than those in ordinary smears but all of them 

 are dried within a few seconds which prevents them from 

 becoming arranged in clusters; also they stick to the 

 glass surface more evenly and tend to form well-spaced 

 groups with all their elements at the same level of focus. 

 Impression preparations from plate cultures have the 

 further advantage of providing ample material during 

 the first few hours of incubation, the stage of the growth 

 cycle when much the clearest results are obtained but 

 when there is very little to scrape off an agar slant. 



To obtain preparations with many cells in the earliest 

 stages of growth a loopful of bacteria from an 1 8-24 hr. 

 slant culture, incubated at 37° C, was emulsified with 

 2-3 ml. of tap water, and three or four drops of this 

 milky suspension were spread on a freshly poured and 

 dried nutrient agar plate. The plate was again dried, 

 incubated and impression preparations were made in the 

 usual manner from small pieces of the medium, cut out 

 and fixed at suitable intervals; an '0 hours' sample was 

 taken immediately the inoculum had dried on the plate. 



Impression preparations were found profitable only 

 if there were not more than one layer of bacteria on the 

 agar. Accordingly plates spread with a few drops from 

 a very rich suspension were used only to study the 

 earliest stages of development and more dilute inocula 

 forming microscopically well-separated colonies, were 

 used for preparations of the growth stages from 4 hr. 

 onwards. Films from confluent growths on solid media 

 were obtained by first making a suspension of the material 

 in tap water, then spreading some of it on fresh agar, 

 fixing it immediately after drying and making an im- 

 pression preparation. 



Particularly clear and instructive preparations of the 

 chromatinic structures of Proteus vulgaris were obtained 



both from the earliest growth stages after subculture on 

 a fresh nutrient medium and from dry smears of fixed 

 material from the raised edge of a colony on an agar 

 plate towards the end of one of the periods of ' consolida- 

 tion' which alternate with periods of rapid swarming 

 (Russ-Muenzer, 1935). Impression preparations of fixed, 

 swarming Proteus filaments show very regular spacing of 

 the chromatinic structures within the filaments, but 

 cytological detail is not very clear. 



B. Staining 



Method 1 . From 70 % alcohol films were transferred 

 directly to N/HC1 warmed to 60° C. and ' hydrolysed ' for 

 about 10 min. They were then rinsed in tap water and 

 two changes of distilled water and floated on the staining 

 solution made with 2-3 drops of Gurr's 'improved 

 Giemsa-stain R66' per ml. of phosphate buffer (Balint, 

 1926), or a mixture of equal parts of tap water and 

 distilled water. Duration of staining differed according 

 to whether the preparations were to be differentiated 

 and dehydrated in the conventional acetone/xylol mix- 

 tures (pure A — A 14 pts, X 6 pts — A 6 pts, X 14 pts — 

 a few seconds in each, followed by three changes of xylol, 

 10 min. in each), or whether they were to be examined 

 undifferentiated, mounted in water. In the latter case it 

 sufficed to stain for half an hour at 37° C, whereas several 

 hours were required when the stained preparations were 

 to be dehydrated and mounted in Canada balsam. In 

 water the outlines of the bacterial cytoplasm are much 

 more clearly defined than in balsam-mounted prepara- 

 tions, the bacteria appear broader, and in hydrolysed 

 preparations cell boundaries are visible together with 

 the chromatinic structures, whereas most of the boun- 

 daries become invisible during the process of dehydration 

 which precedes mounting in balsam. Photomicrographs 

 too are often more satisfactory in their contrasts and 

 richer in detail when taken from water-mounted ma- 

 terial, particularly where very small elements are con- 

 cerned such as the cells of old cultures. Sealed with 

 wax, water-mounted preparations will keep their colour 

 contrasts for several days; finally they may be stained 

 once more, dehydrated and mounted in canada balsam. 



Method 2. A quick way to obtain passably well-defined 

 pictures of chromatinic structures is to dip unfixed, air- 

 dried impression preparations for 5 sec. into boiling N/5 

 HC1, rinse and mount in 0-1 % crystal violet in water. 



Method 3. To stain the cell boundaries, Bouin-fixed 

 films were floated on Giemsa stain of the same strength 

 as that used for nuclear structures, for not more than 

 3 min., rinsed and taken through 70, 96 and 100% 

 alcohol (5-10 sec. in each) to xylol and canada balsam. 

 When examining such preparations an orange light 

 filter was often used. 



Method 4. The Feulgen reaction was carried out ac- 

 cording to the prescriptions given for ordinary cyto- 

 logical work, except in staining B. megatherium for which 

 the optimal conditions of hydrolysis were found to be 

 10-15 min. at 40 instead of 60° C. 



Method 5. To demonstrate the cell wall a plasmolysis- 

 like effect was produced in B. megatherium by dipping 

 air-dried, unfixed vegetative cells grown from spores 

 for 2-2£ hr. under agar strips (see § A (1)) into boiling 

 2-5-8-0% NaOH for 15-30 sec. The films were then 

 rinsed in water and mounted in a 0-5 % watery solution 



173 



