Author´s reply

In further consideration of the analysis of the multivibrator a sign error has been detected. Thus the expression for b is corrected. The error makes b approximately 0.1% greater than actuality. The effect of this error on the results given in Table I is not significant, because of the minor importance of Vco as the source of instability. When the sign error has been corrected and the expression of a, b , and τ are applied to the expression of B in (3) one finds out that B equals zero and only one time constant is then present