Serial dilutions are regularly used in microbio- logical laboratories, mainly to prepare bacterial suspensions suitable for making colony (viable) counts.
A method has recently been described (Hedges, 2002) which allows estimation of the basic (in the sense of ‘irreducible’) variance of such dilutions and of the resultant colony counts. Familiarity with this method and its derivation is assumed in what follows.
The method utilizes only the specifications of the manufacturer to quantify the pipetting error. These comprise a component due to variations in the