shows measurements of total variance (with maximum % RSD of 5 for repetitive measurements) as a function of column length, where the intercept on the y axis corresponds to dispersion caused by the injection valve. The intercept at each flow rate was found to vary from 0.46 to 0.003 min2 , corresponding to 34–23 nL, respectively, with change in flow rate from 0.05 to 0.40