aiorgy.com/ [^3]: I.B.Bondarenko, K.A.Bondarenko, http://www.bondarenko.co.uk/ [^4]: This is still smaller than that observed in many young SNRs (see e.g. fig.2 in @2007ApJ...664.1079E) where the expansion rate is typically $\sim$3.4 to 7 times faster. [^5]: It should be noted that the estimated $E_0$ value is likely to be an upper limit because the $l_0$ for 3C 391 is much greater than in the models used to derive it. [^6]: This is because the shock velocity drops below $1000\;\kms$ before radiative losses become important for this low-density cloud. [^7]: See http://www.sron.nl/divisions/hea/spex [^8]: It should be noted that the SNRs in question could be located closer than their currently estimated distance if they are interacting with an overdensity of gas and the estimated distances are from the far side of the shell. [^9]: This could be because the SNR expands into dense environment resulting in a higher compression ratio, or because of projection effects. [^10]: For example, [@2010arXiv1004.5130C] find that SNRs expanding into a constant density medium should have radio spectral index of $\alpha=-0.65$. For $\alpha=0$ and $\beta=2$ the equation yields an ambient density $n_0 \approx 1.45\;\rm cm^{-3}$. [^11]: The $\gamma$-ray flux of the extended source of HESS J1303-631 is likely to be underestimated since it is not possible to distinguish between the compact source and the extended emission of the SNR. Therefore, the flux values of SNR for HESS J1303-631 include both contributions and should be considered as a lower limit to the true emission. [^12]: http://www.eso.org/instruments/fors/ [^13]: http://space.mit.edu/CXC/software/suzaku/ae\_xis\_20060213.html [^14]: http://pulsar.sternwarte.uni-erlangen.de/wilms/research/tbabs/