Abstract
A vast amount of work has been done which apparently supports the theory that a filtrable stage exists in a life cycle of M. tuberculosis. It should be noted that the outstanding criticism of all such work lies in the fact that in each case use has been made of Chamberland or Berkefeld filter candles which are difficult to standardize inasmuch as they involve factors such as electrical charge of the particle, medium in which the particle is suspended, as well as pore size. Until a filtration process can be devised which would permit exact duplication at any time and which could be standardized and controlled to retain a given size of particle, proof would be inconclusive that a filtrable form of M. tuberculosis exists.
For this reason it was felt that more significant data could be obtained with the ultrafilters of Krueger and Ritter 1 in which such factors as filter charge and adsorption of filtrate play practically no part, the membranes actually operating on the principle of a sieve. By carefully controlled experiments they not only graded the collodion membranes of varying pore sizes on the basis of permeability to colloidal particles of known size but also determined the degree of negative pressure which could be used in each case without causing distortion of the membrane. It was found that by using a negative pressure of 20 cm. a 4.5% collodion membrane (with pore diameter 40μμ) retained all known bacteria.
The following filtration experiments, therefore, were carried out on 4.5% and 5.0% collodion membranes. The strains of tubercle bacilli used had been freshly isolated from sputum and cultivated both on Sweaney's egg-glycerol medium and on glycerine veal broth. Microscopic examination of the culture on Sweaney's medium after 6 weeks incubation revealed many smoothly staining acid-fast rods with few granular forms while in the latter case the acid-fast granules predominated after 3 months incubation.
Get full access to this article
View all access options for this article.