Best known for incredible displays of several thousand meteors per hour on November 27 of 1872 and 1885, today's Andromedids are only weakly represented by displays of less than three meteors per hour around November 14. The radiant is currently located at RA=26 deg, DEC=+37 deg. The associated comet was 3D/Biela, which split and was observed as two comets in 1846 and 1852. The comet most likely produced the 1872 and 1885 displays as a result of a total breakup. The shower's present duration extends from September 25 to December 6, although this is primarily due to the existence of numerous filaments caused by the considerable change the comet's orbit has gone through.
The history of the Andromedids is directly linked to the history of the remarkable comet 3D/Biela. The comet was discovered on three occasions before its periodic nature became known: first by Montaigne (Limoges, France) on March 8, 1772, second by J. L. Pons (Marseilles, France) on November 10, 1805, and finally by W. von Biela (Josephstadt, Germany) on February 27, 1826. The apparitions of 1772 and 1805 involved short observation periods of only 29 and 36 days, respectively, but during 1826, the comet was observed for 72 days, which enabled Biela to mathematically link all three apparitions and declare the discovery of a new periodic comet. The comet was successfully recovered by John Herschel on September 24, 1832.
Comet 3D/Biela was missed at the unfavorable return of 1839, but was recovered by F. de Vico (Rome, Italy) on November 26, 1845. Although a few observations were made in the next month, interest in the comet increased following M. F. Maury's (Washington, DC) January 13, 1846 observation of two distinct nuclei. Observers reported the nuclei to slowly move away from one another and by the end of March they were separated by 14 arc minutes. However, later investigations revealed the increasing separation was due to the comet's steady approach to Earth and, in truth, the nuclei had remained about 1.6 million miles apart during the entire apparition.
Periodic comet 3D/Biela was next observed during 1852. A. Secchi (Rome, Italy) recovered the main comet on August 26, but it was September 15 before the first observations of the secondary comet were made. The somewhat unfavorable approach of the comet caused it to enter the Sun's glare at the end of September, and no observations were made following the 29th. As it turned out, 1852 marked the last time observations were made of comet 3D/Biela. It was poorly placed for observation in 1859, and several months of extensive, but unsuccessful searches during the very favorable return of 1865-1866 caused astronomers to theorize that the comet had completely broken up.
As the story on comet 3D/Biela slowly unfolded, the astronomical world was also becoming aware of a new meteor shower. On the evening of December 6, 1798, H. W. Brandes (Göttingen, Germany) witnessed a large display of shooting stars. He said, "I first noticed them soon after the close of evening twilight, and having no other business, I kept count of the number which appeared in the small segment of the heavens which I could with convenience survey from my seat." His counting revealed rates of about 100 per hour for four straight hours, after which activity dropped off drastically. Brandes noted that occasional glances to other parts of the sky revealed similar quantities of meteors and he estimated "many thousand shooting stars must have been visible above my horizon." The display ended nearly two hours prior to midnight. The only unfortunate aspect of Brandes' observation was the failure to note the radiation point of the meteors. Such was also true on December 7, 1830, when Abbe Raillard (France) recorded the appearance of "many" meteors, but failed to give further details. Fortunately, these two earlier displays were apparently confirmed on December 7, 1838, when observers on the east coast of the United States provided details of a very strong display.
From New Haven, Connecticut, E. C. Herrick, C. P. Bush, A. B. Haile, J. D. Whitney and B. Silliman, Jr., observed during December 6-15, 1838 and noted meteors falling at a rate of 28 to 62 per hour on the evening of December 7. It was noted that many "large and splendid fireballs...attended with trains" were visible on both December 6 and 7. Herrick uncovered additional observations from Connecticut, New York and Georgia, and concluded that the meteors seemed to radiate "not far from Cassiopeia; or perhaps, more nearly, from the vicinity of the cluster in the sword of Perseus" at an overall rate of between 125 and 175 per hour. The Andromedids were next observed on December 6, 1847, when E. Heis (Germany) observed and plotted several meteors from RA=21 deg, DEC=+54 deg [he had originally estimated the radiant as RA=25 deg, DEC=+40 deg, but later revised it---Author].
The 1860's were especially important for the field of meteoritics as G. V. Schiaparelli's recognition of comet Swift-Tuttle's production of the Perseids inspired other astronomers to seek additional comet-meteor associations. Early in 1867, E. Weiss (Austria), H. L. d'Arrest (Germany), and J. G. Galle (Berlin, Germany) independently noted that meteor activity observed in early December of 1798 and 1838 moved in the same orbit as comet Biela. With this link being established, 3D/Biela became one of the first comets to be recognized as a meteor shower producer. Weiss and Galle contemplated a return of the Andromedids in 1872, while d'Arrest predicted a reappearance of activity on December 6, 1878.
Weiss continued to investigate the link between Biela and the Andromedids during 1868. Taking the known orbits of comet Biela for 1772, 1826 and 1852, he noted the comet's ascending node was gradually decreasing, thus causing the theoretical dates of maximum to be December 10, December 4 and November 28, respectively. The resulting radiants would have been RA=18.7 deg, DEC=+58.1 deg in 1772, RA=22.8 deg, DEC=+47.7 deg in 1826, and RA=23.4 deg, DEC=+43.0 deg in 1852. Weiss summarized his paper by predicting that activity might be observed from this radiant on November 28 of 1872 or 1879. Meanwhile, although not being immediately revealed until several years later, Giuseppe Zezioli (Italy) observed fairly weak activity from the Andromeda-Cassiopeia border on November 30, 1867. Overall, seven meteors were plotted and a radiant of RA=17 deg, DEC=+48 deg was revealed. This was the first activity noted from this region since 1838. However, although it offered support to Weiss' recognition of the shower's maximum moving into November, this activity was in no way comparable to the activity of 1798 and 1838.
Comet 3D/Biela was next predicted to arrive at perihelion in 1872, but the few searches made revealed no trace of either component. However, shortly after sunset on November 27, the pulverized remains of 3D/Biela began striking Earth's atmosphere. P. F. Denza (Moncalieri, Italy) and three others observed about 33,400 meteors during 6 hours 30 minutes. Around 8 p.m. (November 27.79 UT) he said the display "seemed a real rain of fire," when meteors fell at a rate of 400 every minute and a half. J. F. Anderson (Pau, France) obtained excellent meteor counts which revealed rates of 30 per minute shortly after 6:30 p.m., which slowly increased to 36 per minute around 7:45 p.m. (November 27.78 UT) and declined to about 14 per minute by 10:30 p.m.
One of the most complete of the 1872 observations came from Stonyhurst Observatory. Being aware of Weiss' prediction of possible enhanced activity, S. J. Perry began observing after darkness had settled. As soon as he had detected notable activity he directed two assistants to aid in watching the sky. The result was a fairly accurate determination of the radiant position as RA=26.6 deg, DEC=+43.8 deg. It was estimated that maximum had occurred around 8:10 p.m. (November 27.84 UT) when meteors were falling at rates too numerous to count. During the 13 minutes prior to 9:00 p.m. one observer counted 512 meteors, for a rate of about 40 per minute. Perry estimated that a total sky rate would then have been about 100 per minute. The character of the shower's meteors was well established by the Stonyhurst observers as they indicated 90% of the meteors were very faint. Perry said a typical bright Andromedid had the appearance of "a white star with a greenish-blue trail." A peculiar feature of the shower involved the simultaneous appearance of meteors which moved parallel to one another. For example, Perry pointed out that at 9:16 p.m. "five burst out close to Gamma Andromedae and travelled eastward together."
Although Western Europe was definitely the best place to be for the maximum of the Andromedids in 1872, observations were also made elsewhere in the world. Most notable were the observations by Hubert A. Newton and others in the eastern portion of the United States. Activity was first noted from the Gamma Andromedae region on November 24 when three-fourths of the 40-50 meteors seen each hour radiated from that star. On the 25th hourly rates were 20-25, with about half radiating from near Gamma Andromedae. Overcast skies were present on the 26th, but the storm was well observed on the 27th. Newton said a party of 2-6 observers counted 1000 meteors between 6:38 and 7:34 p.m. (about November 28.0 UT), with the quantity dropping to 750 in the next hour and twenty-five minutes. The meteors were described as slower than the Leonids and generally faint. Newton and A. C. Twining placed the radiant "in the line from the Pleiades to Gamma Andromedae and 3 degrees beyond that star" (the Author computed this as RA=26 deg, DEC=+44 deg). Newton and Twining described the radiant as longer in RA than in DEC, with the length not being less than 8 deg.
In the years immediately following 1872, activity was totally absent from the region of Andromeda. D'Arrest's predicted 1878 appearance of the Andromedids never took place, nor did Weiss' predicted 1879 return. Shortly thereafter, several astronomers predicted the shower would reoccur on November 27, 1885, and a last-minute reminder was published by Crawford (Dun Echt Observatory) a couple of weeks prior to this date.
As the sun set on November 27, 1885, observers immediately became aware of exceptional activity in the sky. J. Smieton (Broughty Ferry, Scotland) first began observations at 5:30 p.m. and noted meteors falling at a rate of 25 per minute. By 6:00 p.m. (November 27.75 UT) rates had gradually increased to 100 per minute. Something curious occurred at 6:20 p.m., when "a marked decrease in the intensity of the shower was noted." Thereafter, Smieton noted a steady increase to a peak of 70 per minute around 6:38 p.m., after which the shower steadily declined. The radiant was determined as RA=21 deg, DEC=+44 deg. He described the activity as consisting mainly of "shooting stars," but a large number of meteors "had brilliant phosphorescent trains, which continued to glow for several seconds after the meteors themselves had vanished. Occasionally one of the trains would break up into fragments, and in one instance a curious spiral form was assumed."
W. F. Denning (Bristol, England) actually noted activity from the Andromedid region 24 hours earlier, when rates averaged 100 per hour. But on the evening of the 27th, he declared "meteors were falling so thickly as the night advanced that it became almost impossible to enumerate them." He said observers with especially clear skies had rates of about one meteor every second or 3600 every hour.
Additional details of the 1885 Andromedid activity were revealed in the early portion of an 18-page paper written by Newton and published in the American Journal of Science in June 1886. It appears that while some observers experiencing clear skies could not accurately count the meteors visible each minute, others gave quite consistent estimates. At Marseilles Observatory (France), E. J. M. Stephan, A. L. N. Borrelly and J. E. Coggia independently made several counts near the shower's maximum and said the single observer rate reached 233 per minute. Observers at Palermo obtained a similar estimate of 213 per minute during one 5-minute interval. Using these observations, as well as others made in Beirut and Moncalieri, Newton determined the maximum hourly rate as 75,000 under very clear skies. His indicated time of maximum corresponds to November 27.76 UT.
Although Newton's study produced an excellent view of the 1885 Andromedids, he went on to look at the physical characteristics of the stream, as well as how it had evolved. From the 1885 observations, he concluded the stream had an overall thickness of 200,000 miles, while "the really dense portion of the stream was less than 100,000 in breadth." Newton commented on the perturbations the comet had experienced from Jupiter during 1794, 1831 and 1841-1842, and theorized the debris encountered by Earth in 1872 and 1885 must have left the comet after the last encounter with Jupiter, otherwise the perturbations "would have scattered the group, and we should have had a much less brilliant star-shower in 1872 and 1885." He also confirmed Weiss' 1868 discovery of the decreasing ascending node of comet Biela by showing how the actual observations of the shower in 1798, 1838, 1847, 1867, 1872, and 1885 indicated the solar longitude of the shower's maximum had also declined from 256.2 deg to 245.8 deg---meaning the date of maximum had decreased by nearly 11 days.
During the years immediately following 1885, the Andromedids were again nowhere to be seen, but at the next predicted passage of 3D/Biela in 1892 (the comet was of course not located) observers in the United States detected a strong meteor shower. Although it was not of the caliber of the 1872 and 1885 displays, the Andromedids of November 24, 1892 did produce rates of several hundred per hour. In particular, D. Kirkwood (California, USA) observed 150 during one 30-minute interval and reported "an intelligent and trustworthy young gentleman counted 350 meteors in half an hour" later in the evening. Another example comes from C. D. Perrine (Alameda, California, USA), who observed 1013 meteors during one interval of an hour and eighteen minutes.
The Andromedids next reached maximum on November 24, 1899, and on November 21, 1904, with hourly rates of 100 and 20, respectively. These rates, when combined with the longer observed durations of November 23-24 in 1899 and November 16-22 in 1904, indicated the stream was rapidly dispersing. The shower was virtually nonexistent in the years immediately following 1904, but, in 1940, two apparent peaks of activity were noted: an outburst of 30 faint meteors per hour occurred on November 15, according to R. M. Dole (Cape Elizabeth, Maine, USA), while 5 per hour were detected by J. P. M. Prentice (England) during November 27-December 4. These two peaks inspired Prentice to theorize that the Andromedids had divided up into several components.
Although visual activity seemed nonexistent following 1940, some remnants of the Andromedid stream were detected among the over 2000 meteors photographed during the Harvard Meteor Project of the early 1950's. The first official recognition of the photographic Andromedid meteors came in a paper by G. S. Hawkins, R. B. Southworth and F. Stienon published in 1959. Isolating all November meteors detected during 1950-1956 with RA ranging from 0 deg to 50 deg and DEC ranging from 0 deg to 50 deg, they compiled a list of 47 "possible Andromedids." Noting the period of comet 3D/Biela, the authors said associated meteors should have an atmospheric velocity of 20 km/sec and, after sorting out all meteors with velocities between 19 and 21 km/sec, a total of 23 meteors remained. Plotting the meteors by date of appearance, the authors noted a maximum photographic hourly rate of 1.0 was reached on November 14, and they theorized this represented a visual rate of 5 per hour. The duration of activity was given as November 2-22. The authors also examined the duration of the 1872 activity and deduced a thickness of 400,000 miles. They concluded that the failure of activity to appear in 1878 indicated the debris had then spread over less than 4% of the comet's orbit.
A reexamination of the orbit of periodic comet 3D/Biela was conducted by B. G. Marsden and Z. Sekanina during 1971. Subsequently, L. Kresak computed the new encounter conditions between Earth and the comet's orbit and noted that the shower's maximum would occur 12 days earlier than in the past, or on November 17.0, 1971 (essentially confirming the photographic maximum determined by Hawkins, Southworth and Stienon), at which time the radiant would be at RA=26.2 deg, DEC=+24.6 deg (20 deg south of the 19th century positions). Kresak noted that the closest approach of Earth to the comet's orbit was 0.05 AU, while Earth remained within 0.10 AU during November 6 to December 1.
B.-A. Lindblad conducted a computerized stream search during 1971 using 865 precise meteor orbits obtained during the Harvard Meteor Project. On this occasion, Lindblad used the D-criterion calculation and revealed the Andromedids to consist of two streams. Two years later, however, a further investigation by A. F. Cook, Lindblad, Marsden, R. E. McCrosky and A. Posen revealed the 1971 study to have actually placed several members of the Andromedid stream in with the Piscids of September. The subsequent 24 photographic meteors were interpreted as indicating one very complex Andromedid stream, rather than two simple branches. The authors described the existence of "a systematic trend with the longitude of the sun (i.e., with that of the earth) such that the perihelion moves out from the sun, the inclination increases, and the node and argument of perihelion vary in such a way as to keep the longitude of perihelion unchanged." The average orbit based on these 24 meteors is given in the "Orbit" section below. The Author's investigation into the photographic "Andromedids" of Cook et al, reveals the described orbital variations to be real, though it is obvious that the described transition is not smooth. A more direct explanation might be that the present Andromedid stream is composed of numerous filaments---each of which represents a ringlet of material left by Biela during previous evolutionary changes in its orbit.
As has already been pointed out, 3D/Biela underwent several close approaches to Jupiter and it seems likely that debris would have been left in each of the comet's previous orbits. It will be recalled that Weiss said these Jupiter encounters meant a steady decrease in the ascending node. This has continued, so that the date of maximum gradually moved from the second week of December back to mid-November. Comparing the observed orbit of 3D/Biela in 1772, with its hypothetical orbit of 1971 (see the "Orbit" section), it will be noted that perturbations over the last 200 years have also caused a steady decrease in the inclination and perihelion distance, as well as an increase in the argument of perihelion. For meteor observers this indicates that the current Andromedid activity of November comes from the newest orbits, while that of early December comes from the oldest. More importantly this also indicates that from November to December Earth encounters a series of filaments, the orbits of which gradually increase in perihelion distance, inclination, and ascending node, and decrease in the argument of perihelion---the same conditions noted in 1973 by Cook and his fellow researchers.
The 1970's saw a resurgence of interest in the Andromedid (or "Bielid") radiant as several amateur astronomers observed weak activity. During 2 hours on November 22, 1970, M. Hale (Canisteo, New York, USA) detected a rate of 1 Andromedid per hour, while M. Savill (Selsey, England) observed average rates of 4 per hour on November 21/22 and 23/24, 3.5 per hour on November 25/26, 1 per hour on November 26/27, and 2 per hour on December 4/5. During November 12 and 14, 1971, A. Porter (Narragansett, Rhode Island, USA) observed a total of 5 meteors from this radiant, all of which were described as red and of negative magnitudes. Porter commented, "This shower suffers from an inattention it does not deserve, because many amateurs hear somewhere that it's dead."
Observers in England also observed the Andromedids in 1971. R. Mackenzie, director of the British Meteor Society (BMS), said visual observers detected a maximum rate of 3-10 meteors per hour, while "BMS radio observations indicate a burst of faint meteors giving a ZHR of 35 meteors/hour." Mackenzie went on to describe annual Andromedid observations by the BMS extending from 1972 to 1975, in which maximum visual hourly rates attained 10, 2, 4, and 8, respectively.
The most extensive recent observations of this stream comes from the Western Australia Meteor Section. J. Wood encouraged a survey of the Andromedid shower in 1979. Activity was noted during November 10-29 and a maximum ZHR of 3.73+/-1.86 came on November 27 from a radiant of RA=28 deg, DEC=+38 deg. A total of 114 man hours were accumulated, but only 26 Andromedid meteors were observed. The average magnitude of these meteors was 3.42, while 3.8% left trains. The actual hourly rate of the activity peaked at 2 on November 26/27 and 1 on November 17/18. Curiously, Wood's group compiled 76 man hours of observing during November 13-30, 1981 and revealed only 3 Andromedids.
The Author concludes by stressing that the current Andromedid stream is a conglomeration of ringlets formed by the strong perturbations caused by repeated close approaches of comet 3D/Biela and its subsequent debris to Jupiter. Weiss and Newton provided the best evidence supporting the comet's orbital evolution, while observers of the last 100 years have revealed an increasing annual duration of the shower and the existence of several peaks of activity during the month of November. These peaks do not occur every year, and might indicate that the meteoric matter has still not had enough time to evenly distribute itself around the orbits of many of the ringlets.