Very good question but keep in mind that partial charge concept is a gross oversimplification. The very first sentence on Wikipedia on partial charge Partial charge is not fully correct either. Consider the partial charge as a book-keeping "number". Just like the oxidation state concept, a "partial" charge can be estimated for a diatomic molecule AB, on the basis of electronegativities
$\delta_A = G_A -L_A - B[\chi_A/(\chi_A+\chi_B)]$
where, $\delta_A$ is the partial charge, $G_A$ is the group number, $L_A$ is the unshared electron number on A, B is the number of bonding electrons, and $\chi$ is the electronegativity value proposed by Allen. Allen avoids the term "partial charge" and calls it this number "Lewis-Langmuir Atomic Charges." So don't treat this number as a fraction of an elementary charge.
The following references will help:
Ionization Energies, Electronegativity, Polar Bonds, and Partial Charges
Ronald J. Gillespie, James N. Spencer, and Richard S. Moog, J. Chem. Educ. 1996, 73, 7, 627.
"Lewis-Langmuir atomic charges." Allen, Leland C., Journal of the American Chemical Society 111.25 (1989): 9115-9116.
And finally if you are interested in who came up with this $\delta$ symbol see...
Jensen, W. B. (2009). The Origin of the" Delta" Symbol for Fractional Charges. Journal of Chemical Education, 86(5), 545.
An old 1959 article in the Journal of Chemical Education addresses this idea. See Models for demonstrating electronegativity and "partial charge" R. T. Sanderson, J. Chem. Educ. 1959, 36, 507.
Knowing molecules to be composed of two or more “point” positive
charges in relatively fixed positions, imbedded in a diffuse cloud of
electrons, one can recognize that the assignment of partial charges to
the individual atoms must at best be only a deliberate simplification.
It is indeed an attempt to describe an exceedingly complex system in
terms useful toward reaching a practical understanding of the
molecule.
Furthermore he writes,
All these physical data, when considered together with the properties
of compounds, strongly support the conclusion that the condition of a
combined atom does indeed depend on its environment, which to a large
and usually dominant extent results from electronegativities. One may
call this contribution “partial charge” or by any other name, as long
as its true significance is clearly understood: “Partial charge’ is a
quantitative estimate of the relative extent to which differences in
electronegativity of the atoms before compound formation have altered
the electronic field about the nucleus of an atom in a compound. In
the following discussion it will become apparent that as an index of
the condition of a combined atom, partial charge provides a uniquely
useful and satisfactorily reliable basis for explaining much of
chemistry