A new calibration of Hγ equivalent widths is presented. Unlike Petrie's method, the new calibration does not require an estimate of the luminosity class or spectral type of the star. A brief comparison with other calibrations is made.

A composite HR diagram constructed from the main sequences of three moving clusters (Hyades, α Persei, and the Inner Region of Upper Scorpius) using MK classifications, is presented and discussed.

From a careful study of physical multiple systems, it has been possible to derive differentially absolute magnitudes and intrinsic colours of O stars from the corresponding quantities, relatively well-known, of companions of these stars.

The significance of recent improvements made in the absolute monochromatic flux measurements of the Sun and Vega are discussed with a special emphasis on the absolute calibration of the V-magnitude of stars. The difficulties relating to the determination of vacuum ultraviolet fluxes are outlined and the improvements that can be achieved using synchroton radiation as the fundamental laboratory source are presented. Colour indices derived from OAO-2 observation are given for about 80 early type stars and the (1700 – V) colour is discussed in terms of interstellar reddening and sensitivity to effective temperature. Basic problems requiring further investigation are finally outlined.

Ultraviolet observations of line blanketing in the spectrum variables α Scl and α2 CVn show that the usual procedure of allowing line blocking in model atmospheres has to be revised. It is also suggested that the region around 2900 Å could be safely used for determining effective temperatures.

Absolute intensity measurements made in high altitudes on board of an aircraft are of less accuracy than ground based observations made at locations with favourable atmospheric conditions. As things stand, the rms-errors appear to be ⩾ 3% and ⩽ 2% respectively. Some of the reasons are sketched.

The scale of effective temperatures Teff is based on observed absolute radiation temperatures Tr, which are defined by Planck's radiation law where TAu designs the absolute temperature of the gold point. A relative scale of radiation temperatures can be derived from spectrophotometric comparisons with a standard star. The absolute calibration of the standard star (α Lyr or Sun) demands a careful comparison with a standard radiation source of well known spectral energy distribution (Black Body or Synchrotron). With ground-based observations atmospheric extinction is to be taken into account; with extraterrestrial observations detectors may be used which are absolutely calibrated in a radiation laboratory under space conditions.

A comparison of the theoretical and observed indices has been made for the UBV, u v b y and the Geneva's UBVB1B2V1G systems. In the actual state of the photometric calibrations, coincidence between theoretical and observed indices better than are probably not significant.

Absolute flux densities for a zero magnitude A0V star at the effective wavelengths of the u, v, b, y spectral bands are presented.

The radii of 9 long-period Cepheids have been determined by means of the modificated Wesselink method and some general conclusions have been formulated. The relations R – P and Fv – Mv for stars recognized as fundamental and first overtone pulsators have been established.

Regardless of the degree of elaboration of series of models, just how can they be used for calibration purposes? And how much is this calibration sensitive to the quality of the model theory? These two questions are the basis of our discussion, which covers : I – The general principles of the use of model atmospheres in stellar calibration (1 – The two dimensional classifications; 2 – The use of the total luminosity; 3 – The cases of Vega and Sirius; 4 – The calibration of ST – Teff relation); II – The failures of the two parameters model atmospheres (1 – The observational need for more-than-two-parameters classification; 2 – The abundance of elements, the line formation, and the model atmospheres; 3 – Various sources of unadequacy of models; 4 – Envelopes or shell features; their influence on model-building; 5 – The case of HD 45677. Diagnostic of early-type stars; 6 – Various unexplained spectral features); III – The present state of the model factory (1 – The classical models; 2 – New concepts in the description of a stellar atmosphere; 3 – New approaches in model making; 4 – Conclusions).

Ultraviolet observations of Helium-weak stars carried out on OAO-2 indicate that temperature estimates from the ground based region of the spectrum may be unreliable.

A star showing extreme features characteristic of the Ba II stars has been found in the old open cluster NGC 2420. A new colour-magnitude diagram has been constructed for the cluster, and a distance modulus of 11m.7 is obtained by fitting the unevolved main sequence to the zero-age main sequence. The location of the Ba II star is consistent with its being a disc population star of absolute magnitude Mv = −0m.3, mass 1.4 M⊙ and age 2.4 × 109 yr.

In chapter 1 basic methods are reviewed, and applications and suggestions for future work are presented. In chapter 2 a revision is given of the intrinsic-colour relation in the U, B, V system of hot main-sequence stars. Some temperature-colour relations are discussed in chapter 3, where also a correction formula is given for the effects of interstellar reddening on the effective temperatures of hot main-sequence stars. An empirical mass-luminosity relation is given in chapter 4.

We propose, in several tables, a number of stars selected to be observed both by photometrists, spectroscopists, astrometrists. Properties of selected stars are discussed.

V photoelectric magnitudes are determined on an independent basis. Correlation with V sequence of the UBV system confirms the coincidence of the two definitions. Comparisons with the Catalogue of Bright Stars revealed that more than 40% of BS stars have an apparent V magnitude with a probable error of .

Preliminary results of recent direct comparisons between some early type O, B standards (and also α Lyr) and a terrestrial source previously compared to a black body are given for the ultraviolet range. It appears that the previously-used gradients ɸuv must be corrected on the average by +0.10.