We describe the H i component of the Southern Galactic Plane Survey (SGPS). The SGPS is a large-scale project to image at arcminute resolution the H i spectral line and 21 cm continuum emission in parts of the plane of the Milky Way. The survey covers Galactic longitudes 253 l 358 and latitudes jbj 1N5 (SGPS I), plus a first quadrant extension covering 5 l 20 and jbj 1N5 (SGPS II ). The survey combines data from the Australia Telescope Compact Array and the Parkes Radio Telescope for sensitivity to angular scales ranging from 2 0 to several degrees. The combined data cover 325 deg 2 and have an rms sensitivity of 1.6 K. Here we describe the H i observations and data reduction in detail and present examples from the final data products. The complete data set is publicly available through the Australia Telescope National Facility's H i Surveys archive. This data set provides an unprecedented view of the neutral component of interstellar hydrogen in the inner Milky Way.

The nucleus of the Milky Way is known to harbour regions of intense star formation activ-

We present new Faraday rotation measures ( RMs) for 148 extragalactic radio sources behind the southern Galactic plane (253 l 356 , jbj 1:5 ), and use these data in combination with published data to probe the large-scale structure of the Milky Way's magnetic field. We show that the magnitudes of these RMs oscillate with longitude in a manner that correlates with the locations of the Galactic spiral arms. The observed pattern in RMs requires the presence of at least one large-scale magnetic reversal in the fourth Galactic quadrant, located between the SagittariusCarina and Scutum-Crux spiral arms. To quantitatively compare our measurements to other recent studies, we consider all available extragalactic and pulsar RMs in the region we have surveyed, and jointly fit these data to simple models in which the large-scale field follows the spiral arms. In the best-fitting model, the magnetic field in the fourth Galactic quadrant is directed clockwise in the Sagittarius-Carina spiral arm (as viewed from the north Galactic pole), but is oriented counterclockwise in the Scutum-Crux arm. This contrasts with recent analyses of pulsar RMs alone, in which the fourth-quadrant field was presumed to be directed counterclockwise in the Sagittarius-Carina arm. Also in contrast to recent pulsar RM studies, our joint modeling of pulsar and extragalactic RMs demonstrates that large numbers of large-scale magnetic field reversals are not required to account for observations.