The effective temperature of a star -- providing the essential link between
observed properties and results of theoretical stellar structure and
atmosphere models --
can be directly obtained
from measurements of its angular diameter together with the total
absolute flux at the Earth, integrated over the entire spectrum
The prime source in this field remains the seminal work of
Code et al. (1976), who combined angular diameters measured by the
Narrabri stellar intensity interferometer with UV, visual
and IR observations for 32 hot stars, corrected for interstellar
extinction. Smalley & Dworetsky (1995) repeated their analysis
using more recent observational and theoretical
results, and obtained values on average only one percent lower for stars
with 
25kK.
While this method is termed `direct', the allowance for fluxes in
the unobservable Lyman continuum are taken from
model calculations, and so becomes increasingly reliant on
suitable models for very hot stars.
For O and early B stars a significant fraction of the total energy
is emitted shortward of 
=912Å and obtaining 
from
observed energy distributions requires accurate knowledge
of surface gravities (Abbott & Hummer 1985; Hummer et al. 1988).
Therefore, more sophisticated techniques need to
be employed for such stars. We now turn to indirect techniques used for the temperature
determination in the vast majority of hot stars.