Observing the Perseids

For a short summary of this meteor shower, click here

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.

This meteor shower gets the name "Perseids" because it appears to radiate from the constellation Perseus. An observer in the Northern Hemisphere can start seeing Perseid meteors as early as July 23, when one meteor every hour or so could be visible. During the next three weeks, there is a slow build-up. It is possible to spot five Perseids per hour at the beginning of August and perhaps 15 per hour by August 10. The Perseids rapidly increase to a peak of 50-80 meteors per hour by the night of August 12/13 and then rapidly decline to about 10 per hour by August 15. The last night meteors are likely to be seen from this meteor shower is August 22, when an observer might see a Perseid every hour or so.

For observers in the Southern Hemisphere, the Perseid radiant never climbs above the horizon, which will considerably reduce the number of Perseid meteors you are likely to see. Nevertheless, on the night of maximum, it is possible to see 10-15 meteors per hour coming up from the northern horizon.

There are other, weaker meteor showers going on around the same time as the Perseids, but the Perseids will generally appear to move much faster across the sky than meteors from the other showers. In fact, the Perseids are among the fastest moving meteors we see every year. Another way to know if the meteor you saw was a Perseid is to mentally trace the meteor backwards. If you end up at Perseus then you have probably seen a Perseid meteor! If you are not sure where Perseus is in the sky, the following charts will help you find it from both the Northern Hemisphere and Southern Hemisphere:

Location of the Perseids
For Northern Hemisphere Observers

This represents the view from mid-northern latitudes at about 2:00 a.m. local time around August 12 and 13. The graphic does not represent the view at the time of maximum, but is simply meant to help prospective observers to find the radiant location. The red line across the bottom of the image represents the horizon. (Image produced by the Author using SkyChart III and Adobe Photoshop.)

Location of the Perseids
For Southern Hemisphere Observers

This represents the view from mid-southern latitudes at about 6:00 a.m. local time around August 12 and 13. The graphic does not represent the view at the time of maximum, but is simply meant to help prospective observers to find the radiant location. The red line across the bottom of the image represents the horizon. Although the actual radiant is below the horizon, the stars just above the horizon are those of the constellation Perseus. (Image produced by the Author using SkyChart III and Adobe Photoshop.)


The earliest record of Perseid activity comes from 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, 10th 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 in Italy on August 10th; however, credit for the discovery of the shower's annual appearance is given to Adolphe Quételet (Brussels, Belgium), who, in 1835, reported that there was a shower occurring in August that emanated from the constellation Perseus.

The first observer to provide an hourly count for this shower was E. 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 maximum 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 G. V. Schiaparelli (Italy) revealed a very strong resemblance to periodic comet 109P/Swift-Tuttle, which had been discovered in 1862. This was the first time a meteor shower had been positively identified with a comet and it seems safe to speculate that the high Perseid rates of 1861-1863 were directly due to the appearance of 109P/Swift-Tuttle, which has a period of about 135 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.

During 1973, the astronomer Brian G. Marsden examined the orbit of periodic comet 109P/Swift-Tuttle to determine when it was likely to return. The observations from the 1862 return were not the best and the uncertainty in the orbital period amounted to several years. His best bet was to try and identify a previous return. He found two good options: a comet in 1737 and one in 1750. Marsden chose the 1750 comet as the best candidate for a previous appearance of comet 109P/Swift-Tuttle and predicted the comet would return in 1981. This immediately generated excitement among meteor observers as the potential for enhanced activity unfolded. This excitement seems 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 109P/Swift-Tuttle, comet observers were less enthusiastic as the comet was not recovered. Following 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 between 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 comet 109P/Swift-Tuttle was actually seen 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 more interested in the Perseid display of 1993. Predictions indicated Europe was the place to be during August of 1993. Observers from around the world flocked 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.

The Perseid radiant turns out to be complex. The main radiant is situated near the star Eta Persei, but other radiants appear to be active at the same time. As long ago as 1879, W. F. Denning (England) 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 frequently observed during the twentieth century---especially with telescopic aid. 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 existence 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 exhibit radiants that move nearly parallel to the main radiant.

There is an uneven size distribution within the stream. One very interesting characteristic of the Perseids is that there are times when larger, brighter meteors are much more plentiful than smaller, fainter meteors. In 1953, A. Hruska (Czechoslovakia) found that Perseids were brighter during August 8-12, slightly fainter on August 12/13, and notably fainter by August 14/15. In 1956, Z. Cephecha (Czechoslovakia) found the meteors were brightest on the night of Augsut 6/7 and faintest on the night of August 13/14. A similar pattern has been noted by more recent studies during the 1980s and 1990s. All of the magnitude studies have one thing in common---they point to an irregular mass distribution within the Perseid stream. Some of this is most likely due to the Earth encountering filaments of material representing different that comet Swift-Tuttle has moved in during the last 2000 years.

There is an odd variation from year to year in the number of Perseids exhibiting persistent trains. One of the first astronomers to study this was M. Plavec (Czechoslovakia), who examined 8028 Perseids seen during the period spanning 1933 to 1947. He noted the 45% of Perseids exhibited persistent trains in 1933, while this was value changed to 60% in 1936, 35% in 1945, and 53.3% in 1947. Plavec noted that he could not correlate the variations to sunspot numbers. It could be that this is also tied in to Earth encountering different orbital filaments perviously shed by comet Swift-Tuttle.


More orbits have been calculated for the Perseids than for any other meteor stream, with the first coming during the early 1860s. During the last few decades both photographic and radio-echo methods of observing have enabled precise orbit determinations. The Author has accumulated 102 precise photographic meteor orbits from various lists published in the United States of America, the Soviet Union, and Czechoslovakia, and arrived at the orbit below.

ω Ω 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, while Z. Sekanina determined an orbit from data gathered during the 1969 session of the Radio Meteor Project.

ω Ω i q e a
150.9 136.5 113.1 0.95 0.91 11.11
152.5 139.7 110.2 0.960 0.881 8.040

For the sake of comparison, the orbit of periodic comet 109P/Swift-Tuttle is as follows:

ω Ω i q e a
153.00 139.44 113.43 0.958 0.964 26.32