Recent measurements of sonoluminescence produced by a single, stable, pulsating gas bubble indicate that the spectrum can be modeled by that for a blackbody at a temperature of nearly 40 000 K [R. Hiller, S. J. Putterman, and B. P. Barber, ‘‘Spectrum of synchronous picosecond sonoluminescence,’’ Phys. Rev. Lett.69, 1182 (1992)]. These results are in contrast with earlier measurements of the spectrum which is modeled by electronic transitions of rotational and vibrational bands within specific elements of the host liquid [E. B. Flint and K. S. Suslick, ‘‘The temperature of cavitation,’’ Science253, 1397 (1991)]. It is suggested that the single‐bubble SL observed by Hilleretal. is intrinsically different from the multiple‐bubble SL observed by Flint and Suslick. In the former case, symmetric bubble collapse leads to shock wave formation in the gas; in the latter case, asymmetric bubble collapse leads to liquid jets that penetrate the hot bubble interior and result principally in incandescence of the host liquid.