In his interesting commentary “Gamma-Ray Telescopes Show Origins of Cosmic Rays” (Physics Today, January 2010, page 13), Bert Schwarzschild argues that we have finally found strong evidence that supernova remnants (SNRs) accelerate cosmic rays (CRs) in some external galaxies. However, the facts are not so clear-cut. As Schwarzschild himself notes, “Strictly speaking, the LMC [Large Magellanic Cloud] map — and the starburst-galaxy results — tie the CRs only to starforming regions and not specifically to SN remnants.”

Many objects, not just SNRs, correlate with an increased star formation rate. For example, because gamma-ray bursts, pulsars, and superbubbles (multiple interacting SNRs) all correlate with a high star formation rate, they are possibilities for the source of the cosmic rays in the starburst galaxies NGC 253 and M82 and in the LMC. Given the spatial and temporal density of SNRs in NGC 253, M82, and the LMC, a larger percentage of SNRs would be expected to be in superbubbles rather than be isolated; in our own galaxy, about 80% of SNRs are in superbubbles.

And as I recently argued, 1 the distinction between isolated remnants and superbubbles cannot be brushed off lightly; they are physically distinct entities.

In the case of the LMC, it is more likely that superbubbles are accelerating the cosmic rays than are isolated SNRs, since there is evidence for many superbubbles within the LMC coincident with the gamma-ray emission. 2  

1.
Y.
Butt
,
Nature
460
,
701
(
2009
).
2.
See, for example,
B. C.
Dunne
,
S. D.
Points
,
Y. -H.
Chu
,
Astrophys. J., Suppl. Ser.
136
,
119
(
2001
), 
Y.
Butt
,
A.
Bykov
,
Astrophys. J. Lett.
677
,
L21
(
2008
).