We introduce a model for semiflexible polymer chains based on the inte
gral of an appropriate Gaussian process. The stiffness is characterize
d physically by adding a bending energy. The degree of stiffness in th
e polymer chain is quantified by means of a parameter and as this para
meter tends to infinity, the limiting case reduces to the Brownian mod
el of completely flexible chains studied in earlier work. The calculat
ion of the partition function for the configuration statistical mechan
ics (i.e., the distribution of shapes) of such polymers in elongationa
l now or quadratic potentials is equivalent to the probabilistic probl
em of finding the law of a quadratic functional of the associated Gaus
sian process. An exact formula for the partition function is presented
; however, in practice, this formula is too complicated for most compu
tations. We therefore develop an asymptotic expansion for the partitio
n function in terms of the stiffness parameter and obtain the first-or
der term which gives the first-order deviation from the completely fle
xible case. In addition to the partition function, the method presente
d here can also deal with other quadratic functionals such as the ''st
ochastic area'' associated with two polymer chains.