High resolution studies of early type emission line stars

This thesis presents a study of early type, emission line objects on small angular scales. Spectroastrometry is used to separate the spectra of unresolved Herbig Ae/Be binary systems. The separated spectra allow the mass ratio of the systems to be established. The separation and mass ratio distribution of the systems suggest that they form via disk fragmentation. To test this conclusion, disk and binary position angles are compared to a model in which the circumstellar disk and binary orbits are co-planar. The data are consistent with the coplanar model and thus with the scenario of disk fragmentation. A search for optical outflows associated with Herbig Ae/Be stars is conducted using integral field unit spectroscopy. No outflows are detected. Therefore, the data cannot constrain the location of the proposed transition from collimated to un-collimated outflows, if such a transition exists. Spectroastrometry is used to probe the origin of the Brackett line emission in the data with sub-milli-arcsecond precision. To explore the potential of spectroastrometry to detect circumstellar disks, spectroastrometric observations of Be stars are presented. It is shown that spectroastrometric data with an angular precision of 0.1 milli-arcseconds can detect milli-arcsecond sized disks. Comparing the spectroastrometric signatures to models of disks with different kinematics demonstrates that Be star disks rotate in a Keplerian fashion, which constrains their formation mechanism. Finally, spectroastrometry over CO bandhead emission is employed to search for disks around massive young stellar objects. Fitting the observed overtone profiles with a model of a circumstellar disk allows the spatial distribution of the CO to be predicted. No signatures of disks are detected. This is entirely consistent with the best fitting models of the emission, and thus with the presence of small-scale disks around massive young stellar objects.