# Dispersion in Chromatography Columns - Dispersion Processes that take Place in an LC Column > Longitudinal Diffusion > The Diffusion Process > Page 17

Now, from the Plate Theory (see Plate Theory and Extensions),

_{},

where (n) is the variance of the Gaussian curve.

Now, (n) is
the volume variance of the Gaussian curve (*i.e., *_{}), then, by
comparison, (2D_{m}t) will be the length variance _{} of the
concentration curve where (t) is the elapsed time. Consequently, if a
differential equation of the form _{} is derived that
describes some form of dispersion that arises from a random diffusion process,
then the solution will be a Gaussian function and, more important from the
point of view of the Rate Theory, the Gaussian curve will have a variance given
by (2D_{m}t).