Spectroscopic constraints on outflows from BN-type objects

New high-quality high spectral resolution observations of the HI line emission from massive young stellar objects are described and discussed. It is proposed that two distinct physical components contribute to the observed emission. One of these is an optically-thick high-velocity stellar wind, the other a more slowly moving optically-thin volume of gas that may, in the case of S106IR at least, be caused by mass loading of the stellar wind. This decomposition is shown to resolve a long-standing problem regarding the relative widths of high and low opacity lines.