Bernard has proposed a measure of the effective gluon "mass" in terms of the screening of the linear potential for adjoint sources, and has extracted an estimate of mE=500-800 MeV from a lattice Monte Carlo calculation. The same quantity is calculated in the bag model, with the result mE=740±100 MeV. This arises as an effect of confinement, even though the gluons themselves are massless. The model illustrates several interesting features of this measure of "mass", and these are discussed.