Sawai Jai Singh II's Jantar Mantar in Jaipur, a UNESCO Site Since 2010, Counts 19 Stone Instruments and a Giant Sundial Rated Accurate to Two Seconds

The Jantar Mantar in Jaipur is an 18th-century observatory of lime plaster, marble and a little bronze, and UNESCO inscribed it on the World Heritage List in 2010 under criteria (iii) and (iv). UNESCO’s statement of value calls it the most significant and best preserved set of fixed monumental instruments built in India in the first half of the 18th century, and says some of them are the largest ever built in their categories.

Sawai Jai Singh II, ruler of Amber and founder of Jaipur, had it built for watching the sky with the naked eye. Its largest instrument, the Brihat Samrat Yantra, is a sundial that the ICOMOS evaluation says measures local time with a precision of 2 seconds. Jai Singh ruled from Amber, one of the forts in our piece on the hill forts of Rajasthan, before he moved his capital to the new city beside the observatory.

How many instruments, depending on who counts

Michel Cotte’s case study for ICOMOS (the heritage council that advises UNESCO) and the International Astronomical Union counts 19 main instruments or groups of instruments. BBC Sky at Night Magazine, in a guide by Khursheed Dinshaw, also says 19. ICOMOS’s own evaluation lists eighteen main instruments, mentions nineteen in the nomination dossier’s table, and says these contain 35 individual instruments. UNESCO’s summary says “some twenty”. The spread comes from how paired and grouped instruments are counted.

Most were built of brick rubble and plaster and a few of bronze, according to Cotte. Four have mainly metal structures, ICOMOS says: the Unnathamsa Yantra, Chakra Yantra, Krantivritta II and Yantra Raj. Masonry surfaces are clad in lime mortar, usually red with white details, and the parts used for measuring are fine lime plaster or marble. ICOMOS puts the highest point on the site, the summit of the Brihat Samrat’s gnomon, at 19 metres above ground.

A king who found his brass instruments too small

According to the British Museum’s biography, Jai Singh pointed out to the Mughal emperor Muhammad Shah that astronomical discrepancies possibly affected the dates of Hindu and Muslim holy days, and that he received imperial backing to build observatories at Delhi, Jaipur, Varanasi, Ujjain and Mathura. For the technical trigger, the Astronomical Society of India’s outreach page says observed positions of heavenly bodies did not match the predictions of the traditional Siddhantic calculations.

Saudi Aramco World quotes the preface to his astronomical tables on the choice of stone. Jai Singh first built brass instruments like those of the Samarkand observatory, including an armillary sphere two metres across. He judged them not accurate enough, citing their small size, the lack of divisions into minutes, worn axes and shifting centres, and he decided on large stationary instruments with graduations cut into marble or limestone. He named the tables Zij Muhammad Shahi after his patron, and the British Museum says the work was completed in 1728. A zij is a table of star and planet positions. Jai Singh’s aim, Aramco’s Paul Lunde writes, was to bring Ulugh Beg’s 15th-century star catalogue up to date.

Many instruments are described in the Hindu school of astronomy, ICOMOS notes, yet no trace remains of any Hindu observatory earlier than the 18th-century ones at Jaipur, Varanasi, Delhi and Ujjain. UNESCO places Jaipur in the tradition of Ptolemaic positional astronomy, shared by many civilizations, and credits the observatory with helping complete the Zij tables. Ulugh Beg’s Samarkand observatory, the evaluation adds, influenced both the Mughal and the Jaipur observatories.

Built from the late 1720s, used as the city’s clock

Sources give different spans for the construction. The Ministry of Tourism’s Incredible India site says 1728 to 1734. Work began in the 1720s and finished in 1738, writes Cotte. Construction seems to have peaked in 1734 and 1735, when 23 astronomers worked alongside masons and engravers, and it continued until 1738, according to ICOMOS. In 1734, the evaluation adds, two French Jesuit scholars worked out the latitude and longitude of Jaipur precisely.

A permanent team of around twenty astronomers worked for the maharaja. The observatory set Jaipur’s local time and announced it to the city by drum rolls or cannon fire, ICOMOS says, and its results went into almanacs until recent times. Right up to 1944, Lunde writes, the Jaipur observatory was used to establish correct time. Tradition links the Brihat Samrat Yantra with forecasting the monsoon and the harvest, ICOMOS adds, and the meteorology of the monsoon is in our piece on the science of the monsoon.

After Jai Singh died in 1743, repairs were made at least twice, including in 1771. Astronomical work ended around 1800, ICOMOS says, and a neighbouring cannon foundry then used the site as an annex. Maharaja Ram Singh II’s reign saw the first substantial restoration, completed in 1876, and Lieutenant A. H. Garrett, the British resident engineer at Jaipur, led a major one in 1901 and 1902. Cotte records the most recent campaign as running from 2006 to 2007.

What the main instruments measure

Every reading is made by eye, from shadows or by sighting along a pillar or ring. The Brihat Samrat Yantra is a triangular wall whose sloping top edge is parallel to the Earth’s axis, flanked by two quadrants lying in the plane of the equator. In ICOMOS’s figures the quadrant radius is 15.15 metres, and the wall is 22.6 metres high including 3.5 metres below ground, with a hypotenuse of 50.1 metres. The instrument reads local time from the shadow on the quadrants and the declination of stars at night, and a staircase leads to a small cupola at the top.

Other accounts give larger heights: 27 metres in the Border Sundials description and in a seismic-risk passage elsewhere in the ICOMOS evaluation, 27.5 metres in Sky at Night, and over 27 metres in Aramco. None of these pages reconciles the figures.

Four meridian dials in two chambers make up the Sasthamsa Yantra, graded from 0 to 60 degrees, for the declination and zenith of the Sun or a star. The Jai Prakash Yantra is two hollow hemispheres, each 5.4 metres across in the ICOMOS description and about 5.5 metres in Sky at Night. Together they produce an inverted image of the sky, with a coordinate grid marked on the bowl. A suspended disc casts a shadow that shows the Sun’s path through the zodiac and its position, and a hole in the disc allows night observations. Cotte says the observer can move freely inside the bowl to take readings.

The Great Ram Yantra is a pair of cylindrical structures 4.5 metres high with an inside diameter of 6.95 metres, each with a central gnomon (shadow-casting pillar). It gives the altitude and azimuth, the height and compass bearing, of celestial bodies. The Yantra Raj carries a metal astrolabe 2.43 metres tall, which ICOMOS and Cotte both describe as probably the largest in the world. The Rasivalaya is a set of twelve instruments, one for each sign of the zodiac, with gnomons from 4.2 to 6.2 metres. Two marble bowls make up the Kapala Yantra, which converts positions between the horizon system and the equatorial system, a direct visual transformation in Cotte’s words. Between them, Cotte says, the instruments cover each of the three main classical coordinate systems: the local horizon system, the equatorial system and the ecliptic system.

What the two-second figure rests on

Four sources carry the two-second claim: the ICOMOS evaluation, Cotte’s study, Barry Perlus’s jantarmantar.org and Sky at Night. ICOMOS adds that the State Party, which means India, defined precision levels for thirteen of the eighteen instruments. Four were rated high-precision: the Brihat Samrat, Laghu Samrat, Sasthamsa and Dakshinottara Bhitti Yantras. The rest were rated medium or low.

Perlus and Border Sundials both note a practical limit. Because the Sun is about half a degree across, the shadow’s edge is soft, and Perlus says the blur can cover six or seven of the two-second marks. Readers must estimate the shadow’s centre, and both pages describe a thin twig or taut string held about a centimetre above the scale as a way to sharpen the reading. Border Sundials also says the common description of the Samrat Yantra as the world’s largest sundial is not true, without setting out its reasons on that page.

Lunde, a historian, gives his view that Jai Singh was probably mistaken in believing that larger instruments gave finer readings, because European astronomy advanced by minimising error with the vernier, the micrometer and the telescopic sight. A related point about words comes from ICOMOS, which calls “pre-telescopic” ambiguous for an observatory built more than a century after Galileo, and prefers to describe Jaipur as a late and final flowering of naked-eye observation. India’s reply to ICOMOS was that large fixed instruments in the open air proved sturdier and more precise than bronze ones for observing with the naked eye. India’s current high-altitude research telescopes are covered in our article on Hanle in Ladakh.

The four sister observatories

Delhi’s observatory dates to 1724, built at Muhammad Shah’s request, Lunde writes, and has four main instruments, the Samrat, Ram, Jai Prakas and Misra Yantras. Misra means “mixed”, because that instrument combines several devices. In his preface Jai Singh claims to have invented the Jai Prakas, Ram and Samrat instruments himself, which Lunde says may well be true, while noting he may also have had descriptions of similar instruments used at Samarkand.

At Ujjain, the district administration’s page dates the observatory to 1719 and says Jai Singh built it while governor of Malwa. That is earlier than Lunde’s 1724 for Delhi, and published accounts of the order differ. The page lists a sundial, the Narivalaya, the Digansha and a transit instrument as Jai Singh’s, and says the Digansh Yantra was rebuilt in 1974 and the Shanku Yantra in 1982. The page adds that the observatory has recorded daily weather observations and has published an annual ephemeris, the Panchang, since 1942. Incredible India says some instruments there are still used for research.

ICOMOS calls the Varanasi observatory the Man Singh Observatory. The Astronomical Society of India’s outreach page says instruments at Jaipur and Ujjain have measurement markings in reasonable condition, while those at Delhi and Varanasi are in disrepair, with restoration under way at Delhi. The same page lists observatories in four cities only: Delhi, Jaipur, Ujjain and Varanasi. Mathura, named as the fifth by the British Museum and by Lunde, is not among them. Jai Singh built the three outer sites, Lunde writes, so that readings from one could be checked against another.

India told ICOMOS in 2010 that it intended a national serial nomination of the group of four at Jaipur, Delhi, Varanasi and Ujjain, to be followed later by a wider international one. ICOMOS replied that the strategy was not for it to judge at that stage.

What the restorations changed

The Committee’s 2010 decision lists requests to the State Party, among them a management plan, a single overseeing authority for the site and its buffer zone, and a review of how earlier restorations altered the graduated scales. The concern, as ICOMOS explains it, is that the scales were originally cut in lime plaster, sometimes filled with lead, and many were rebuilt in engraved marble from the 19th century. A restoration in 1901 converted the original graduations to Western hours, minutes and seconds, and nothing is known for certain today of the original markings. The scientific function of the great majority of the large instruments has nonetheless been retained, ICOMOS concludes.

Cotte writes that the Department of Archaeology and Museums of Rajasthan manages the site. UNESCO records that the site is protected under the Rajasthan Monuments, Archaeological Sites and Antiquities Act of 1961 and was designated a monument of national importance in 1968. In its 2010 evaluation ICOMOS reports, citing the State Party, that visitor numbers averaged 3,500 a day, with peaks of 10,000.