The galactic magnetic field is commonly supposed to be due to a dynamo acting
on some large scale seed field. A major difficulty with this idea is that
estimates of reasonable seed field strengths tend to be quite low, on the order
of $\sim10^{-20}$ gauss. Here we examine the contribution due to the flux
entrained in winds from protostars formed in the first dynamo e-folding time of
a galaxy's existence. Using a minimal estimate of a protostellar magnetic field
we find that if each protostar ejects a single current ring, sufficient to
maintain flux freezing in the wind, than the large scale average dipole field
from all such current rings will be at least 5 orders of magnitude larger than
previous seed field estimates. Allowing for a reasonable amount of magnetic
activity in protostars during an extended period of mass loss increases this to
a dipole seed field of $\sim10^{-12}$ gauss. For the purposes of producing a
seed field it is irrelevant whether or not this initial injection of flux takes
place in a newly formed galactic disk, or in star forming proto-galactic
clouds. The compression of this dipole field into a thin disk will lead to a
large scale $B_r\sim 10^{-10.5}$ gauss. Initially, field strengths on smaller
scales will be larger, but nowhere near current levels.