408 MYCOBACTERIUM 



it was often impossible to find acid-fast bacilli in films, even though the presence of tubercle 

 bacilli could be shown by culture and by pathogenicity experiments, he devised a number 

 of difierent staining methods to determine whether bacilli of any sort could be demon- 

 strated microscopically. The method that he found most successful was to stain for 

 24 to 48 hours in aniline gentian violet or carbol methyl violet at a temperature of 37° C, 

 to treat with Lugol's iodine solution, and to decolorize by a mixture of absolute alcohol 

 and clove oil, or by a dilute mineral acid, and a mixture of alcohol and acetone. Examining 

 some mihary tubercles of a calf that had been injected with virulent bovine bacUh, he 

 failed to find any baciUi in preparations stained with Ziehl-Neelsen ; but in preparations 

 stained by his own method he found large numbers of fine rods in the tubercles, often 

 accompanied by small rounded granules arranged singly, in pairs, or in short chains 

 resembUng beaded bacUU. The rods preponderated in the necrotic portions of the tubercles, 

 the granules in the peripheral zones ; both were coloured violet, and both were numerous 

 within the cells. Small pieces of the lung were seeded on to the serum slopes, and incu- 

 bated at 37° C. ; smears were examined daily. In smears stained with Ziehl-Neelsen no 

 baciUi were found for 6 days, when acid-fast rods appeared ; but in smears stained by 

 Much's method fine granules and rods were visible after 3 days. Small pieces of the lung 

 injected subcutaneously into guinea-pigs gave rise to generahzed fatal tuberculosis in 

 8 weeks ; acid-fast baciUi were found in the tissues of the dead animals. Much obtained 

 similar results with other tuberculous material. He concluded that — ^(i) There is a form 

 of tubercle bacillus that is not stainable by Ziehl, but is stainable by Much's method ; 

 it is granular, (ii) In tuberculous organs this granular form may be the only stainable 

 form of bacillus present, (iii) The granular form may be accompanied by fine rods, which 

 likewise do not stain with Ziehl. (iv) The granular forms are virulent, (v) There are 

 transition forms between the Gram + granules, the fine Gram + rods, and the acid-fast 

 rods and granules. 



For a long time comparatively little attention was paid to Much's work, but of late 

 years a number of observers have studied the growth of acid-fast bacilli in suitable culture 

 preparations, and have demonstrated the presence of granular forms similar to those 

 described by Much. The interpretation of these forms, however, has given rise to con- 

 troversy. While Sweany (1928) and Kahn (1930) hold that they represent a stage in the 

 life-cycle of the baciUi, Oerskov (1932) beheves that they are products of degeneration. 

 The bundles of extremely fine rods that Kahn described as forming part of the life-cycle 

 are interpreted by Oerskov as crystals, formed partly from the medium and partly from 

 baciUary products. According to Yegian and Porter (1944), many of the non-acid-fast 

 forms are frank artefacts resulting from trauma to the organisms during the preparation 

 of the film. These observers find that destruction of the integrity of the cell is accom- 

 panied by a loss of the acid-fast staining property. Yoimg organisms are more readily 

 destroyed than old, and the greater the effort spent in spreading the growth with a spatula 

 on the slide, the more numerous are the non-acid-fast rods and granules seen microscopioally. 

 The micro-motion pictures obtained by Wyckoflf (1934) and WyckofF and Smithburn 

 (1933) show that the young baciUi increase in size before dividing, but that, as the culture 

 ages, division contiiuies without previous enlargement. The resulting organisms, there- 

 fore, become shorter and shorter, tiU true coccoid forms, staining intensely acid-fast, 

 appear. Transplanted into a fresh medium, these short forms again give rise to typical 

 baciUi. The sequence of events is so similar in general outline to the behaviour of non- 

 acid-fast bacteria, that there seems no justification for postulating the existence of any 

 special cycle of development. 



Filtrable Forms of the Ttiherde Bacillus. — Closely connected with the presence of Much 

 granules is the existence of the so-caUed filtrable forms of the tubercle baciUus. Since 

 Pontes' (1910) original observation, numerous workers have claimed to demonstrate 

 under appropriate conditions the presence of filter-passing forms possessing a low degree 

 of pathogenicity for guinea-pigs and constituting a special stage in the life-cycle of the 

 organism. This subject was reviewed in the second edition of Topley and Wilson (pp. 



