The GEOS symposium held on 7–8 February 1976 in Marly‑le‑Roi gathered 35 observers to review the group’s activities after a year marked by more than 140,000 measurements and the launch of SIGMA. Participants assessed ongoing observing programmes and early analytical work, including the development of computer‑based processing for periodic and non‑periodic variables. It was agreed that partial results would circulate internally, while synthesis studies would be published in SIGMA. Verified discoveries could be submitted to the IBVS.
A major result presented was the discovery of RL 1 CVn’s variability, made possible through the identification of a new effect called “Rollandisation,” described in the document as oscillations caused when a comparison star is itself variable (“si on mesure une variable au moyen de deux repères dont l’un est lui-même variable…”). This led to the detection of similar cases and to the exploration of shorter variability periods. Rapid‑cadence observations on stars such as 1 Per raised the question of distinguishing true stellar flickering from atmospheric scintillation.
The symposium highlighted significant gaps in knowledge concerning many variable‑star classes, including Algol‑type binaries, Cepheid stars, T Tauri stars, δ Scuti stars, and even well‑known objects like CY Aqr, for which only 105 maxima had been published in 50 years. Discussions also addressed observer motivation and the need to focus on measurement quality rather than quantity.
To reduce dispersion, the 1976 observing programme was streamlined: only a few Mira stars (T Cep, χ Cyg, R Leo) were retained, flare stars were removed due to professional coverage, and P Cyg was kept because its constancy remains uncertain. The meeting strengthened coordination within GEOS and aimed for fewer but more scientifically valuable observations in 1976.