Comments on "A note on the Routh-Hurwitz criterion"

In a recent note, it was pointed out that the "ε method" in the application of Routh-Hurwitz criterion fails when the characteristic equation has an arbitrarily small coefficient In this correspondence it is shown that the cause of such a failure is the wrong interpretation of the arbitrariness of ε.