The idea in this note is to use the tensor Taylor series to reproduce the experimental value of the Lamb shift as in Eq. (18) and to find the vacuum fluctuations needed to give the Lamb shift. This is done on the classical level without use of the Dirac delta function.