1. Ultraviolet spectra (1400–1800 Å) of Ap, Am and normal A stars are needed by F. Praderie, R. Bonnet and R. Cayrel.

The spectral resolution has to be nearly 1 Å. Accurate relative photometry (5%) and absolute calibration (30–50%) are required.

A rocket experiment, proposed to ESRO by M. Combes and P. Felenbok is planned for launch in 1972.

2. As neutral silicon and magnesium are very efficient ultra-violet absorbents, A stars ultraviolet fluxes are very faint (Praderie, 1968).

Then a very luminous optical set-up and a high efficiency receiver have to be used. A 30 cm in diameter concave objective grating is associated with a Lallemand electronic camera. The grating (2000 //mm; //l) is holographically made (Labeyrie, 1969). The electronic camera is electrostatically focussed. A semi-transparent solar-blind CsL photocathode is used (Carruthers, 1966).

3. A little mirror, placed against the grating and forming a direct view of the sky, permits to establish an absolute wavelength scale.

During the fly, before and after stellar observations, a little concave mirror mounted into the opening side-door is used to form on the photocathode a spectrum of a Deuterium calibrated lamp. Two photomultipliers, one on each side of the electronic camera, control the lamp stability.

The complete mounting is calibrated in the laboratory using a thermopile as reference, before the launch and after the recovery of the waterproof payload.

4. The chosen stars are the brightest Ap and Am stars: α Dra (Ap; mv = 3.64; equivalent type A 0) and α2 Lib (Am; mv = 2.75; equivalent type A3-A7).

It seems to be possible to obtain spectra (1400-1800 Å) of the Ap star with a spectral resolution of 1 Å and a signal to noise ratio better than 40. But at a pinch one may accept a resolution of 2 Å and a signal to noise ratio of 15 for the shortest range of the Ap star spectrum.