Sunspots from the past

  Pierre Gassendi (1592–1655) is the French priest and astronomer. His six-volume work is known as Opera Omnia. Rare sunspot observations were published in the fourth volume, Astronomica, together with astrometric measurements of the Sun, the Moon, Mars, Venus, Jupiter, Saturn, etc. In 1658, the book was published in Lyon (Lugduni) by Anisson and Devenet, and in 1727, a second version was republished in Florence (Florentiæ) by Celsitudinis. The engraving of drawings in these editions differ, which is noticeable for large sunspots. The orientation of the images does not always coincide, which at a glance indicates that the orientation of the drawings in the book is not always the same as might be seen in the telescope. Opera Omnia also contains drawings of the eclipse, and these are also different in the two editions.

   1633 – 1638

  The sunspot observations were made by means of a Galilean telescope ("Telescopio") in a room with closed windows and doors ("ianua, fenestrisque aliunde occlusis"). The image of the Sun was projected onto a sheet of paper (van Helden 1976). The depicted circle probably had a diameter of 10 inches or about 25 cm ("decussis pedis Parisiensis", which could be translated as ten small units of the Paris foot). The diameter was divided into 120 equal parts (or subunits). For several sunspots, Gassendi gave their width and distance to the limb, disc centre, and ecliptic ("Eclipticae") or diameter ("Diametri") in subunits. Note that the transit of Mercury on 7 November 1631 observed in Paris was projected onto a circle with a diameter of three-quarters of a Paris foot ("dodranti pedis Parisiensis"), that is, 9 inches or about 23 cm, divided into 60 subunits.

  Here, we would like to warmly thank Roger Ceragioli (professional optician and specialist in the history of telescopes at the University of Arizona Steward Observatory Mirror Lab) and Albert Van Helden (professor of history, Rice University, The Galileo Project), who kindly helped us with the Latin texts.

  We sorted 57 spots into 17 groups. In Figure 1a, our group counts slightly differ from those of Hoyt and Schatten (1998). Vaquero et al. (2011) eliminated one spurious observation by Gassendi on 1 December 1638 from the group sunspot numbers. This record does not appear in Astronomica. Another data point that also should be removed is the observation on 7 November 1631, when the transit of Mercury had been reported. Gassendi thought that it was a spot that he had not noticed on the Sun on the previous day. Later he realised that he was indeed observing the planet (van Helden 1976, for more details).

  Figure 1b shows the areas of the individual sunspot groups from the book published in 1658 in Lyon; the largest is Group G17 on 1 November 1638 with 5958 msh. For the first observations, Gassendi gave a size and position of sunspots in subunits of the solar disc. We compared these values in the text with the sizes of spots in drawings of the Lyon book and found that they are in good agreement. If Gassendi indeed observed powerful sunspot populations, it might indicate that the time of Gassendi's (1630s) as well as Galilei's (early 1610s) and Scheiner's (1611 – 1631) observations fall on a secular maximum of the solar activity. For instance, areas of individual groups exceeding 4000 msh were observed during Cycle 18 (Figure 1c, data from the RGO/USAF/NOAA catalogue). The smallest group is G5 on 9 April 1633 with 388 msh. Given that groups smaller than 400 msh make up about 90% of all groups in the Greenwich catalogue, the Gassendi observations may lack a significant portion of sunspots. On the other hand, Gassendi showed sunspots in a schematic manner; neither umbra nor penumbra were depicted. Therefore, the sunspot areas have a significant uncertainty.

  To demonstrate the discrepancy between historical drawings and the RGO/USAF/NOAA catalogue, we compared their distributions. Figure 2 shows the distributions of sunspot group area for Gassendi, for Galilei from Vokhmyanin and Zolotova (2018a), and modern observations. The majority of the Gassendi drawings covers almost half a year from October 1634 to February 1635. From the modern observations, we chose three half-year periods with the largest medians of the sunspot group area distribution (June – November 1874, May – October 1889, and May – October 1957). The distribution for Galilei's drawings is consistent with modern data, while the median of Gassendi's distribution is shifted to larger sunspot groups. Moreover, the occurrence frequency of medium and large groups exceeds that for modern data. Apparently, the size of the groups in Gassendi's drawings is enlarged.

  Figure 3a compares the two large groups G11 (4286 msh) and G17 (5958 msh) from Gassendi's drawings as published in 1658 in Lyon. For comparison, we overlay active regions AR338 from the Kislovodsk Solar Station and AR14886 from the Greenwich photoheliographic catalogue (i.e. the largest active region of NOAA AR14886). The position of Group AR14886 is shown reversed from right to left limb.

  We found that the drawings of Opera Omnia published in Lyon in 1658 differ from those published in Florence in 1727. Sunspot positions and areas were extracted from Gassendi (1658). Figure 3b compares the images for 10 April 1633 from the two books. The different inclination of the diameters indicates that the orientation of the solar disc in Opera Omnia might not be the same as was seen in the telescope. The sunspot latitudes in the Lyon and Florence images diverge by up to 3^o (Figure 3c). A group composed of three spots that are bound together has an area of about 4100 msh in the Lyon book and about 8000 msh in the Florence book. This shows that sunspot-group areas may differ by a factor of two.

  The problem of uniformity of a sunspot-group area as measured at various observatories has not been solved so far. For instance, according to the Kislovodsk Solar Station, AR338 has an area of 3416 msh (on 20 October 2014) and 4096 msh (on the next day); according to RGO/USAF/NOAA, this active region is defined as AR12192 with 2180 msh (on 20 October) and 2410 msh (on the next day).

  The series of sunspot observations on 1–12 April 1633 was made by Gassendi in Diniæ or Diniâ (today Digne-les-Bains), from 25 October 1634 to 28 February 1635 in Aquas-Sextias or Aquis-Sextijs (today Aix-en-Provence), and since 15 October 1635 again in Diniæ. The style of the drawings is not uniform. In the first series of observations from 1–12 April 1633, Gassendi usually drew a new solar image for each subsequent day. Since 25 October 1634, he presented trajectories of sunspots across the disc over several days.

  On 1–12 April 1633, the position of the solar diameter was fixed at around 9 o'clock in the morning. However, the scatter of sunspots on the disc indicates that either they were drawn at different times during the day and/or the images were rotated. On 13 April 1633, Gassendi observed the Sun at 7, 8, 9, and 10 o'clock and claimed that there were no more traces of sunspots ("sed nullum amplius Maculæ vestigium"). We recall that Gassendi reported large spots, hence this statement does not necessarily indicate vanishing sunspot activity at the end of a series of observations.

  On 25 October 1634, within the first noon hour ("sub horam quidem primam à meridie"), Gassendi discovered a double ("gemellam") sunspot, the first for a long time, as he stated. At about the same hour on the next day, Gassendi saw this double sunspot accompanied by a smaller spot ahead and two spots behind. At noon of 27 October, the same objects were vaguely visible. At noon of 28 October, the double sunspot was visible, but the front and two trailing spots had disappeared. Subsequent observations were also made at noon.

  Around or after noon ("sub meridiem") on 24 November 1634, the sunspot observation was made by secretary Antonius Agarratus, who also assisted Gassendi in his astrometric measurements. He saw a double spot ("Maculam geminam detexit") that lived up to the first day of December. Several smaller sunspots are depicted between two large ones in the drawings of this group. Subsequent observations were not always made precisely at noon ("quòd non fuerit semper exquisitè meridies"). After the sunspot group went beyond the limb, Gassendi wrote that on days 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 14, and 15 December there were no sunspots.

  On 20 December 1634 and on subsequent days of this series, the observations were made in the afternoon. From 11 to 17 January 1635, Gassendi did not indicate the time of sunspot observations. From 7 to 10 February 1635, Gassendi presented astrometric and weather reports. He briefly noted that on 10 February and on the following days, sunspots were observed. On 27 and 28 February 1635, the time of the sunspot observations is unknown. On 15 October 1635, a triple sunspot appeared. The sunspot was also observed on 19 October. Gassendi reported spotless days from 24 October to 27 October and from 29 October to 2 November. On 30 October 1638, after the altitude of the Sun reached 26^o, Gassendi wrote that near the vertical of the solar image, to the east and down, sunspots appeared. On the next day, these sunspots were observed in the morning and in the evening. On 1 November, the observation time is unknown.

  This was a partial translation of Gassendi's reports on sunspots. The time of the observations is not clearly stated or mentioned. We assume that on 8–12 April 1633, Gassendi did not observe the sunspots at 9 o'clock; the images could also have been rotated prior to publication (Figure 3b). These findings cause us to ignore the time stamps and instead seek for the right angle of the image at which the spot latitudes would be constant from day to day. Sunspots of each individual day were transferred to a separate solar disc. Then, by rotating images (Method II, see Heliocoordinates), we minimised the discrepancy (we define this value as FF ) between the sunspot latitudes.

  For each day, the coloured bars in Figure 4 show FF  whose minimum values (black points joined