Comparative Counts of Bacterial Nuclei Made Visible by Four Different Techniques

Abstract

Four different methods were used to make visible the nuclei of Escherichia coli strains B and WP 2, and Bacillus subtilis. They were: (1) Giemsa staining after acid hydrolysis; (2) fluorescence microscopy after acridine-orange treatment; (3) acriflavine staining and phase-contrast microscopy; (4) phase-contrast observations with high refractive index mounting medium (poly-vinylpyrrolidone). Statistical analyses showed small but significant differences in nuclear counts between the methods. Methods 1 and 2 were somewhat preferable to methods 3 and 4 in that 1 and 2 were easier to score and there was less difference between observers. It was calculated that a single mean nuclear count, by any observer or method, based on 200 scored bacteria should 19 times out of 20 estimate the "true" mean nuclear number of E. coli WP 2 to within about[plus or minus]7.5% and of B. subtilis and E. coli B to within about [plus or minus]12%. By increasing the number of bacteria counted and the number of methods used these limits could be appreciably decreased.