As a summary of the principal results presented at the ESRO symposium on Calibration Methods in the Vacuum Ultra Violet (Munich, 1968) a description is given of three different absolute intensity calibration methods. These are the branching ratio method, the synchrotron radiation method and the black-body radiation method, and they define the present state of the art.

At the Institute of Physics and Astronomy of the Academy of Sciences of the Estonian S.S.R. there is a laboratory for the research on the properties of detectors for vacuum ultraviolet radiation. The laboratory has also the necessary measuring equipment, including a beam double vacuum monochromator for the wavelengths 1000–2700 Å with a balancing-out system of registration attached to it. The last-named device automatically eliminates effects on the results of measurement caused by the distribution of energy in the spectrum of the light source as well as by absorption in the apparatus.

Earlier the absolute efficiency of detectors was determined by means of a thermocouple but at present it is calculated from the data obtained from the photoionisation cross sections of gases.

The present report throws light on some results of the investigation of the properties of radiation detectors under different operating conditions.

A four inch hermetically sealed reflecting telescope has been attached to the prime-focus scanner to provide an observing system for absolute calibration work. Light sources, including a platinum and a copper furnace, are observed at a distance of approximately 1200 feet. Since the telescope focal length is only 13 inches these sources look like stars. Thus bright stars such as α Lyr are compared directly with the light sources under identical instrumental conditions.

It has been found that horizontal extinction can be abnormally large; therefore it is measured several times each night. The absolute calibration is being derived at present from 3500 Å to 11000 Å and will be extended to 3200 Å as soon as possible. The calibration lies approximately midway between that measured recently by Hayes and that adopted some years ago by Oke. A very preliminary value for the absolute flux from α Lyr at 5556 Å is 3.5 × 10−20 ergs sec−1 cm−2 Hz−1 or 3.4 × 10−9 ergs sec−1 cm−2 Å−1.

Hall's and Stebbins and Whitford's extinction pair has been used to determine interstellar extinction in the ultraviolet. The extinction shows a feature suggestive of graphite. Previous observations of Stecher (1965) are confirmed. The results are presented in the figures.

We present a preliminary analysis of a number of spectrophotometric scans of reddened and unreddened early-type stars obtained with the OAO-A2 satellite. The principal findings are: (1) all the extinction curves show a ‘bump’ at 1/λ ≈ 4.5; (2) most extinction curves show a minimum at 1/λ ≈ 5.5–6.0; (3) all the extinction curves show a rapid rise in the far ultraviolet; and (4) there appear to be real differences in the far ultraviolet extinction. A brief comparison of our observed extinction curves is made with theoretical particle models.

Silicate, ice, and silicate core-ice mantle particles are considered with a view to describing or predicting ultraviolet features in both extinction and polarization which depend on the dielectric nature of the particles.

Mixtures of graphite particles ejected from carbon stars and silicates from oxygen-rich giants are capable of producing excellent fits to the observed interstellar extinction curve. The fits extend over the entire wavelength range of the observations, and include a hump at ~ 2200 Å in the ultraviolet due to the graphite components of the mixtures. The agreement with the observed extinction curve remains good if the small silicate particles acquire mantles of either ice or solid H2. If no mantles are present around the silicate grains comparable mass densities of graphite and silicates are indicated; if mantles of either ice or solid H2 are present the mass density of silicates may be an order of magnitude below that of graphite in the interstellar medium.

Usually selective interstellar extinction is derived by comparing the spectrum of a reddened star with an unreddened nearby star, preferably of the same spectral type. But in many cases this comparison is not possible, so that new methods have to be used.

A direct measure of this quantity is possible when a spectrum shows a number of emission lines as for the peculiar stars (novae, nova-like objects, symbiotic stars, etc.) and for nebulae. From a statistical point of view there are a small number of objects but they are of very high interest as far as stellar evolution is concerned.

Interstellar extinction curves (mext vs. 1/λ) of 20 Å resolution have been obtained at the DAO from photoelectric scanner observations in the range 4000 Å to 5000 Å for five stars, and of 50 Å resolution for four stars in the range 4000 Å to 6500 Å from Willstrop's photoelectric data. There is a closely linear section between 4900 Å and 5800 Å for all of the curves. There are changes of gradient or discontinuities associated with the broadest diffuse interstellar bands at 6180 Å, 4882 Å, 4761 Å and 4430 Å. There is a marked discontinuity near 5800 Å and for some stars a broad absorption near 4200 Å. The 4430 Å band lies between two unequal wings of anomalously low extinction (one of which has been detected at Edinburgh). The irregularities vary from star to star, and those in the neighbourhood of the 4430 Å band seem to have the same form as those in the region of the absorption peak at 2200 Å

Effective temperatures of O-type stars imbedded in diffuse nebulae are derived from measurements of Hα and radio fluxes from the nebulae and the apparent magnitudes of the stars. Accurate model atmospheres, with ultraviolet line blanketing where appropriate, are used for the theoretical relation between effective temperature and the ratio of Lyman continuum to visual stellar fluxes. Although there is considerable scatter in the results, an average temperature of 48000 K is found for spectral type O5, 40000 K for O6, and 35000 K for O7.

Silicon and especially carbon bound-free absorptions considerably reduce the emergent flux in the ultraviolet for stars near spectral type A0 (Teff = 10000 K). An overabundance of silicon and/or an underabundance of carbon can affect the Balmer discontinuity and the Paschen continuum by a few per cent at most. However, the abundances of these ultraviolet absorbers will have little effect on the temperature distribution calculated for a star if the model is chosen to match the visual spectrum. An examination of the ultraviolet spectrum of Sirius shows that still more opacity is needed; part of this absorption can be supplied by line blanketing.

This paper first briefly describes model atmosphere grids now available for comparison with observations. The recent recalibration of the absolute energy distribution of α Lyr substantially improves the agreement of models and observations in the visual. Temperature scales determined by various methods agree reasonably well except for the hottest stars. Recent ultraviolet results suggest that earlier observations of O- and B-type stars indicating large flux deficiencies were probably in error. However, late B- and A-type stars may emit less energy in the UV than that predicted by models which do not include the opacities caused by silicon, magnesium, and carbon.

An absolute spectrophotometric calibration of Vega and eleven other standard stars was completed by the author in 1967. This calibration extends from 3200 to 10870 Å, and therefore gives a useful calibration of the Paschen discontinuity as well as the Balmer discontinuity, for which calibrations have been available for some time.

The measured values of the ratio of the size of the Paschen discontinuity to the size of the Balmer discontinuity have been compared with corresponding values predicted by model atmospheres. The comparison shows that LTE models satisfactorily predict the continuum in the wavelength range given above.

A stellar temperature scale has been determined by fitting the measured sizes of the Balmer discontinuity of 43 stars to blanketed model atmospheres. This temperature scale agrees very well with Morton's modified form of the Morton-Adams temperature scale for all except the latest B-stars, for which higher temperatures are predicted.

Observed ultraviolet fluxes have been compared with the predictions of line blanketed model atmospheres for stars of spectral type later than B5. The comparison reveals significant differences which increase towards later spectral type and which lead to differences between empirical and theoretical bolometric corrections. Empirical bolometric corrections for the spectral range B8–F5 have been computed from a combination of ultraviolet, visual and infrared observational data. These have been used to derive empirically based effective temperatures for five A and F type stars for which angular diameter measurements are available.

Photometric data in the 1050–1180 Å and 1230–1350 Å wavelength ranges, and electronographic spectra in the 1000–1600 Å wavelength range, were obtained in an Aerobee rocket flight on January 30, 1969. The spectral intensities derived from these data for main-sequence stars are in good agreement with the model atmospheres of Morton and co-workers. Giant and supergiant stars, however, appear to be up to one magnitude weaker, at 1115 Å, than main-sequence stars of the same spectral class.

The correction for interstellar reddening appears to be not inconsistent with a 1/λ extrapolation of earlier determinations of Smith (1967) and Stecher (1965), except in the case of θ Ori, in which the predicted color excess appears to be much too great, confirming the existence of a peculiar reddening law in the Orion Nebula region.

We have used the television photometers in the Celescope OAO experiment to measure the far ultraviolet brightnesses of several thousand stars, including parts of the constellations Draco, Lyra, Puppis, Vela, Taurus, and Orion; and the Moon. As of this date (22 July 1969), three of our four cameras continue to operate satisfactorily, and we are making measurements in additional star fields distributed throughout the sky. Our shortest wavelength band, which includes the Lyman α line of atomic hydrogen, provides information on the Earth's outer atmosphere, as well as on star brightnesses. The intensity of the Lyman α radiation from the geocorona is a maximum when the Sun is near the horizon as seen by the OAO, and a minimum when the Sun is in the nadir. The direction that the telescope points does not much affect the measured intensities.

Because of the heavy logistic problems of identification, calibration, and measurement for so many stars and because of the survey character of the program, the scientific interpretation of the results is, as expected, lagging the measurement program. However, one consistent picture beginning to emerge from our data is that our observed stars are about equally divided between those that fall within 0.5 magnitude of the predicted ultraviolet brightnesses and those that are significantly fainter than predicted. Most of the giant stars we observe exhibit these ultraviolet deficiencies. Since some of these giants are stars for which little or no interstellar reddening is predicted, we attribute the observed deficiencies to the stars themselves.

Many of the objects we observe do not have accurate ground-based published data regarding magnitude, color, and spectral type; new ground-based observations of these objects are required to ensure satisfactory interpretation of our results.