The distinction between viruses and plasmids is further blurred
by the habit of some viruses of converting into plasmid form
upon entry to the cell. For example, Eptein-Barr virus will
circularise in the cell to form a stable plasmid that replicates
in phase with the cell. Expression of latency-associated proteins
allows the virus to persist without lysing the cell.
Circularisation of DNA is also not sufficient to distinguish a
plasmid from a virus, since some viruses, for example baculoviruses
(Nucleopolyhedrovirus, Granulovirus), circoviruses (Circovirus) and
corticoviruses (Corticovirus) have circular genomes.
The ability of a virus to package its DNA (or RNA) into a protein
capsid would be a feature that no plasmid shares; in other words,
plasmid DNA outside the cell would be "naked". However, given the
complex means that bacteria have developed for transferring plasmids
between themselves, I would not be surprised if some bacteria have
developed packaging systems for plasmids. However, this packaging
would not have been coded by the plasmid. Also, it is possible for
some viruses to integrate into the genome of the host cells and be
transmitted vertically from a parent to its progeny without the
need for expression of viral structural proteins. The retrovirus
mouse mammary tumour virus is such an example.