Hi, There is a curious history behind the erroneous date for the start of daylight saving in France in 1923 as listed in the current issues of the Annuaire du Bureau des Longitudes.
In the 1924 issue, the date is given as 27 May (no hour cited). https://gallica.bnf.fr/ark:/12148/bpt6k6571678p/f265.item This is also confirmed in a short note by T. (= Émile Touchet?) in the Bulletin de la Société Astronomique de France of the same year. https://articles.adsabs.harvard.edu/pdf/1923LAstr..37..277M Then for several years (1925 - 1934) there is no "heure d'été" table. In the 1935 issue the "heure d'été" table appears again https://gallica.bnf.fr/ark:/12148/bpt6k6573371c/f212.item but with the date changed to 31 March (at 23h) and a reference to a law which was only published nearly two months later in the Journal officiel de la République française https://gallica.bnf.fr/ark:/12148/bpt6k6476007v/f4.item Apparently, when the editors decided to reinstate an "heure d'été" table for the issue of 1935, they inadvertently applied this rule to the current year while the actual start of daylight saving was not until nearly two months later on the day when the law was made public. I have not checked all issues yet but this error in the Annuaire appears to have remained unnoticed up until 1960 https://gallica.bnf.fr/ark:/12148/bpt6k9651481s/f206.item The 1961 issue https://gallica.bnf.fr/ark:/12148/bpt6k9656689b/f261.item finally has the correct date as 26 May (23h). However, this date was inexplicably changed back again to the wrong date in the Annuaire of 2002 https://gallica.bnf.fr/ark:/12148/bpt6k9645974b/f66.item and this error appears to have been unchallenged ever since. I will notify the editors of the Annuaire of this error and hopefully they will have the correct value reinstated in next year's edition. Rob van Gent
