Be stars were discovered in 1867 by Father Angelo Secchi who observed emission lines in Cassiopeia - the archetypal Be star. They are different to normal dwarf or giant B stars because they exhibit, or have at one point during their lifetimes, Balmer emission lines in their spectra. This emission is attributed to a circumstellar disc whose creation remains enigmatic, yet seems to be linked with the Be stars’ high rotation velocities.

Observations of Be stars are complicated as the circumstellar disc obscures and or contaminates the radiation received from the star. Therefore methods must be found to separate the observed flux which is formed from three distinct processes and/or quantities; (i) the intrinsic radiation from the photosphere of the star, which is augmented by (ii) radiation from the hot circumstellar disc, formed through free-free and free-bound interactions, (iii) both these quantities are then changed by their travel through the interstellar medium and wavelength specific scattering (i.e., reddened in the interstellar medium).

The methods to perform this separation are developed in this thesis and then exploited to investigate Be stars. In order to develop these techniques a multi-wavelength observational study of a representative sample of Be stars is carried out to collect photometric data. This thesis presents new IR data (JHK) and optical Strömgren photometry (uvby) which is used in conjunction with existing spectral data and a review of previous theoretical and observational literature encompassing the subject.

The basic photometric data reduction to be performed on the Be star observations is exactly the same as the data reduction performed on all photometric data. The procedure must be repeated hundreds of times for the Be data alone and with the astronomical community generating an ever increasing amount of data there is a necessity for an autonomous procedure. The Robotic Liverpool Telescope will produce vast amounts of data each night. The automated nature of the observations can be exploited to create a bespoke data pipe-line which will perform all necessary data reduction on site.

The process of automating data reduction is referred to as pipe-lining, this thesis develops a pipe-line to reduce data from the Liverpool Telescope. This pipe-line generates reliable and reproducible results with associated errors, thereby removing the need to store intermediate data steps, reducing data storage needs, and greatly increasing the speed of obtaining results.

In order to demonstrate the effectiveness of the pipe-line, thereby establishing its worth, it is run on two distinct data sets. These consist of (i) the Be data previously mentioned and (ii) a set of low mass star data. Low mass star data has been chosen as the reciprocal data set as they are faint (sky-limited) as opposed to the bright Be stars. These data sets therefore represent two extreme boundary conditions on which to test the pipe-line. It is demonstrated that autonomous pipe-line reduction can yield results equivalent to those of manual reduction in a much reduced time-frame.

Once the Be star data has been reduced by the pipe-line, the a procedure devised by Fabregat and Torrejon (1998) is applied to the Strömgren photometry in an attempt to disentangle the circumstellar and interstellar reddenings. It is shown that the large uncertainties associated with this technique, caused by the very small value of the excess at optical wavelengths makes the technique unreliable. Instead a new technique based on the JHK infrared photometry is derived and shown to be more successful.

Significant correlations between the infrared excess emission from the disc and emission from a range of lines (H, Br, Br11 and Br18) are detected. Also found is a significant correlation between v sin(i) (and also sin(i)) of the Be star and the infrared disc emission. These correlations place constraints on the future models of Be star disc structure and formation.