Original material by Gary W. Kronk. Copyright ©1988, 1995
Original graphics by Eric S. Young. Copyright ©1995

Observer's Synopsis

This meteor shower is generally visible between July 23 and August 22. Maximum occurs during August 12/13 (Solar Longitude=139 deg), with the radiant located at RA=48 deg, DECL=+57 deg. The hourly rate typically reaches 80, although some years have been a s low as 4 and as high as 200. The meteors tend to be very fast, possess an average magnitude of 2.3 and about 45% leave persistent trains. The radiant advances by a rate of 1.40 deg/day in RA and 0.25 deg/day in DECL.


This is the most famous of all meteor showers. It never fails to provide an impressive display and, due to its summertime appearance, it tends to provide the majority of meteors seen by non-astronomy enthusiasts.

The earliest record of its activity appears in the Chinese annals, where it is said that in 36 AD "more than 100 meteors flew thither in the morning." Numerous references appear in Chinese, Japanese and Korean records throughout the 8th, 9th, 10 th and 11th centuries, but only sporadic references are found between the 12th and 19th centuries, inclusive. Nevertheless, August has long had a reputation for an abundance of meteors. The Perseids have been referred to as the "tears of St. Lawrence ", since meteors seemed to be in abundance during the festival of that saint on August 10th, but credit for the discovery of the shower's annual appearance is given to Quételet (Brussels), who, in 1835, reported that there was a shower occurri ng in August that emanated from the constellation Perseus.

The first observer to provide an hourly count for this shower was Eduard Heis (Münster), who found a maximum rate of 160 meteors per hour in 1839. Observations by Heis and other observers around the world continued almost annually thereafter, with ma ximum rates typically falling between 37 and 88 per hour through 1858. Interestingly, the rates jumped to between 78 and 102 in 1861, according to estimates by four different observers, and, in 1863, three observers reported rates of 109 to 215 per hour. Although rates were still somewhat high in 1864, generally "normal" rates persisted throughout the remainder of the 19th-century.

Computations of the orbit of the Perseids between 1864 and 1866 by Giovanni Virginio Schiaparelli (1835-1910) revealed a very strong resemblance to periodic comet Swift-Tuttle (1862 III). This was the first time a meteor shower had been positively identif ied with a comet and it seems safe to speculate that the high Perseid rates of 1861-1863 were directly due to the appearance of Swift-Tuttle, which has a period of about 120 years. Multiple returns of the comet would be responsible for the distribution of the meteors throughout the orbit, but meteors should be denser in the region closest to the comet, so that meteor activity should increase when the comet is near perihelion (as has been demonstrated by the June Boötids, Draconids and Leonids).

As the 20th-century began, the maximum annual hourly rates of the Perseids seemed to be declining. Although rates were above Denning's derived average rate of 50 per hour during five years between 1901 and 1910, the observed rate in 1911 was only 4 and fo r 1912 it was 12. Denning wondered whether the shower was declining, but hourly rates seemed to return to "normal" in the years that followed. Quite unexpectedly the shower suddenly exploded in 1920, when rates were estimated to be as high as 20 0 per hour. This was extremely unusual as it came at a time when the parent comet was nearing aphelion! Although a few weaker-than-normal years occurred during the 1920's, the Perseids regained their consistency thereafter, and, except for abnormally high rates of 160 and 189 during 1931 and 1945, respectively, nothing unusual was observed up through 1960.

During 1973, Brian G. Marsden predicted Comet Swift-Tuttle would arrive at perihelion on September 16.9, 1981 (+/-1.0 years). This immediately generated excitement among meteor observers as the potential for enhanced activity unfolded. This excitement see ms to have been fully justified, as the average rate of 65 per hour during 1966-1975 suddenly jumped to over 90 per hour during 1976-1983---with the high being 187 in the latter year. Although meteor observers seemed content with their observations of the enhanced activity from Swift-Tuttle, comet observers were less enthusiastic as the comet was never recovered.

Since the 1983 peak, hourly rates for the Perseids declined. With a full moon occurring just a day before maximum in 1984, the Dutch Meteor Society still reported unexpectedly high rates of 60 meteors per hour. In 1985, reported rates generally fell betwe en 40 and 60 meteors per hour in dark skies, and results were generally the same in 1986.

As the 1990s dawned, Marsden published a new prediction. If P/Swift-Tuttle was actually the same comet seen by Kegler in 1737, then the comet might pass perihelion during December 1992. The comet was recovered late in the summer of 1992. Although not one of the most spectacular apparitions, the comet was well observed. But meteor observers were waiting for the Perseid display of 1993. Predictions indicated Europe was the place to be during the Perseid maximum of 1993. Observers from around the world flock ed into central Europe and were met with hourly rates of 200 to 500. High rates were still present during 1994, this time with the peak occurring over the United States.

From the 1860s onward, studies of the Perseids began to include more than just hourly rates. Numerous observers began to plot the paths of meteors onto star charts to derive the points from which the meteors seemed to be radiating. The most prolific obser ver of this stream was William F. Denning, who, between 1869 and 1898, observed 2409 Perseids and became the first person to derive a daily ephemeris of the radiant's movement. In 1901, he published his most precise radiant ephemeris as follows:

Perseid Radiant Ephemeris

Date RA (deg) DECL (deg)
July 27 27.1 +53.2
July 29 29.3 +53.8
July 31 31.6 +54.4
Aug. 2 33.9 +55.0
Aug. 4 36.4 +55.5
Aug. 6 38.9 +56.0
Aug. 8 41.5 +56.5
Aug. 10 44.3 +56.9
Aug. 12 47.1 +57.3
Aug. 14 50.0 +57.7
Aug. 16 52.9 +58.0

[A recent plotting of 102 precise photographic meteor orbits by the author supports the general accuracy of the above ephemeris with the daily motion of the radiant being computed as RA=+1.40 deg, DECL=+0.25 deg]

In addition to this main radiant near Eta Persei, there have been indications that several secondary showers are also active. Minor activity near the main Perseid radiant has been noted on several occasions up to the present time and may have been noted a s long ago as 1879, when Denning pointed out that he had "detected the existence of two other simultaneous showers from Chi and Gamma Persei." This latter shower is one of the most active of the secondary radiants and seems to have been frequent ly observed during the twentieth century---especially with telescopic aid. The following observations represent some of the details.

  • In 1921, Ernst Opik (Dorpat) observed the Perseids telescopically on August 10 and 12. On the latter date, he noted that 9 meteors emanated from an oval area 5.7 degx2.2 deg across centered at RA=40.0 deg DECL=+55.6 deg.
  • On August 11, 1921, C. P. Adamson (Wimborne, Dorset) seemed to have detected both the normal radiant and this southern component visually by claiming the Perseids emanated from an elongated area extending from RA=43 deg, DECL=+57 deg to RA=49 deg, DEC L=+58 deg.
  • On August 10, 1931, C. B. Ford and B. C. Darling found a telescopic radiant at RA=40.9 deg, DECL=+54.4 deg.
  • On August 8, 1932, Öpik found a radiant at RA=39 deg, DECL=+54 deg.
  • During August 12-13, 1934, Ford found a telescopic radiant at RA=43.1 deg, DECL=+55.2 deg.

    One of the most recent examples of the complexity of the Perseid meteor shower was revealed in three studies of the radiant conducted during 1969 to 1971, by observers in the Crimea. In addition to the main radiant near Eta Persei, they confirmed the exis tence of the major radiants near Chi and Gamma Persei, as well as minor radiants near Alpha and Beta Persei. These meteor showers are generally short-lived and possess radiants that move nearly parallel to the main radiant. The following are summaries of the most consistent of the secondary Perseid radiants.,

  • The Gamma Perseids mainly occur during August 11 to 16 from an average radiant of RA=41 deg, DECL=+55 deg. The radiant diameter averages about 2 deg. Rates rise and decline with those of main radiant.
  • The Chi Perseids occur during August 7 to 16 from an average radiant of RA=35 deg, DECL=+56 deg. The radiant diameter is about 2 deg. Maximum seems to occur between the 9th and 11th.
  • The Alpha Perseids occur during August 7 to 24 from an average radiant of RA=51 deg, DECL=+50 deg. The radiant diameter averages about 1.5 deg. Maximum seems to occur somewhere between the 12th and 17th.
  • The Beta Perseids occur during August 12 to 18 from an average radiant of RA=47 deg, DECL=+40 deg. The radiant diameter averages about 1 deg. Rates are irregular. Weakest branch of the Perseid cluster.

    These secondary centers of activity have been predominantly visual displays; however, time was taken to seek out some of these other radiants during the Jodrell Bank radio-echo survey of the 1950's. Only the Alpha Perseids were noted with confidence. Detected in both 1951 and 1953, the radiant was very diffuse and 8 deg in diameter centered at RA=54 deg, DECL=+48 deg. It was detected between August 8 and 11, and the highest radio-echo rate reached 37 per hour (the main Perseid radiant reached radio-ec ho rates of 50 per hour during the same years).

    Other studies conducted by amateur and professional astronomers during the last 30 to 40 years have involved specific details of shower members. One especially interesting statistic that has been brought forward was the trend that the Perseids seem to be brighter before the date of maximum than afterward. In 1953, A. Hruska (Czechoslovakia) found the average magnitude to be about 2.5 during August 8 to 12. However, on August 12/13 it had dropped to 2.8 and by August 14/15 it had fallen to 3.4. In 1956, Z denek Ceplecha also showed a similar, though less pronounced decline in brightness. During August 4 to 10, the average Perseid was near magnitude 2.68, while during August 10 to 15 it was 2.94. The extremes came on August 6/7 (magnitude 2.31) and August 1 3/14 (magnitude 3.18). Just as Hruska and Ceplecha's studies show conflicting patterns representing the decline in the Perseid magnitude distribution during August, two very recent studies seem to support both views.

    During 1983, members of the Spanish astronomical group Agrupacion Astronomica Albireo, under the direction of Eduardo Martinez Moya, obtained an excellent series of Perseid magnitude observations, which seemed to support Hruska's study. Between August 1 a nd 13, 1983, the average daily magnitude varied from 1.75 to 2.04. Thereafter, it dropped to 2.19 by the 14th, 2.52 by the 15th, 2.77 by the 17th, 2.92 by the 19th and 3.45 by the 20th. Robert Mackenzie (director of the British Meteor Society) claims the magnitude distribution of the Perseids "gives an indication of the particle mass variation in the cross-section of the stream encountered by the Earth." This variation seems to support Hruska's study.

    Another excellent series of magnitude estimates were made by Paul Roggemans (Brussels, Belgium) during July 27 to August 16, 1986. Observing in darker skies than the Spanish group, Roggemans detected 1315 Perseids and gave the average magnitude of the sho wer as 3.10. Roggemans' estimates were very consistent throughout the shower's duration with variations being typically less than 10% on any given day. However, there were two exceptions. The first came on August 5/6 and 6/7, when the average magnitude dr opped to a low of 3.54. The second drop occurred on August 9/10 and 10/11 when the average magnitude reached 3.71. This set of observations seems to support Ceplecha's study.

    All of the above magnitude studies (and many more not discussed here) seem to have one thing in common---they point to an irregular mass distribution within the Perseid stream. Filamentary structure seems the best explanation. During some years, the filam ents are encountered in rapid succession by Earth's passage through the Perseid stream, thus accounting for the consistent magnitude estimates followed by a steady decline. In other years, the filaments are spread out across the stream's width, thus causi ng the consistent average magnitude estimates to be disrupted by periods of activity from primarily brighter or fainter meteors.

    Another statistic that has been brought forward during the last 30 to 40 years has been the percentage of Perseids that possess persistent trains. This is a major factor long noted in the separation of Perseids from other active showers occurring during t he first half of August. Miroslav Plavec used the records made at the Skalnate Pleso Observatory (Czechoslovakia) to produce one of the most ambitious studies of train phenomena to date. He studied 8,028 meteors observed between 1933 and 1947, and found t he following percentages: 45% possessed trains in 1933, 60% in 1936, 35% in 1945 and 53.5% in 1947. The variations could not be correlated to sunspot numbers. Taking an average of meteor train activity noted in various publications between 1931 and 1985, the author has found the average value to be 45% for nearly 60,000 meteors.


    More orbits have been computed for the Perseids than for any other meteor stream, with the first coming during the early 1860's. During the last few decades photographic and radio-echo techniques have enabled the first precise orbital determinations. The Author has accumulated 102 precise photographic meteor orbits from the various United States, Soviet Union and Czechoslovakian lists. The following orbit was revealed.

    AOP AN i q e a
    149.2 139.5 113.2 0.942 0.902 9.641

    Two major radar surveys revealed the Perseids during the 1960's. B. L. Kashcheyev and V. N. Lebedinets found the Perseids during a 1960 survey at the Kharkov Polytechnical Institute (BL1967), while Zdenek Sekanina determined an orbit from data gathered du ring the 1969 session of the Radio Meteor Project (S1976).

    AOP AN i q e a
    1960 150.9 136.5 113.1 0.95 0.91 11.11
    1969 152.5 139.7 110 .2 0.960 0.881 8.040

    The orbit of periodic comet Swift-Tuttle (1862 III) is as follows:

    AOP AN i q e a
    153.00 139. 44 113.43 0.9582 0.9636 26.3169

    Original material by Gary W. Kronk. Copyright ©1988, 1995
    Original graphics by Eric S. Young. Copyright ©1995