On the stability of a linear dipolar fluid