Kodaikanal Solar Observatory Has Recorded the Sun Since 1904, and in January 1909 Its Spectra Revealed the Evershed Effect

The Kodaikanal Solar Observatory sits 2,343 metres up in the Palani hills of Tamil Nadu, and the Indian Institute of Astrophysics (IIA), which runs it, says its white-light pictures of the Sun have been taken with a 6-inch telescope since 1904. The institute describes the result as one of the longest continuous series of solar data. In January 1909 a spectrograph at the same site recorded sunspot gas moving outward across the spot’s outer ring, a motion John Evershed wrote had not, to his knowledge, been observed before. It is called the Evershed effect.

The observatory was founded on 1 April 1899, and in 2025 India Post put its two-domed main building on a commemorative stamp.

A committee, a hill and a transfer from Madras

The idea came from Madras (now Chennai). IIA’s history page says that the government astronomer there, Norman Pogson, proposed in 1882 that photography and spectroscopy of the Sun and stars needed a 20-inch telescope at a hill station in South India. Pogson died in 1891 and Charles Michie Smith took charge of the Madras Observatory.

In 1893 a high-level committee chaired by Lord Kelvin recommended a solar observatory at Kodaikanal, chosen for its southern, dust-free, high-altitude site, according to the same page. Equipment and staff began moving from Madras in 1895, and the observatory was founded on 1 April 1899. Regular solar observations started in the early 1900s.

The instruments that started the record

The long series rests on two instruments. IIA’s instruments page says a small refractor was brought up from Madras in 1898 for visual work on the Sun, and that direct photography of the Sun began in August 1903. The telescope was later fitted with a 6-inch lens, and the new 6-inch version began operating on 31 July 1912. The same page says the shutter that exposes each picture for about a thousandth of a second still works on the same principle for the daily photographs.

Deccan Herald, reporting the stamp release, gives a different history for the 6-inch telescope: installed at Madras in 1850, remodelled in 1898 and set up at Kodaikanal in 1900. The two accounts agree that the instrument came from Madras.

The second instrument is a spectroheliograph, which builds a picture of the Sun in one narrow colour of light. IIA’s history page says it was ordered from the Cambridge Scientific Instrument Company in 1902 and arrived in 1904, built to a design by George Ellery Hale, who invented the type. Set to the Ca II K line (the violet absorption line of ionised calcium, at 393.37 nanometres), it shows the bright patches called plages that mark magnetic activity. With a different arrangement, the same instrument photographed the Sun in the red hydrogen-alpha line at 656.3 nanometres. IIA says its data run from 1904 to 2007.

IIA’s data page lists the start years of the three digitised series: white light from 1904, Ca II K from 1907 and hydrogen-alpha from 1912. The history page adds that nearly 60,000 prominences, the glowing arcs of gas at the Sun’s edge, were recorded between 1904 and 1914.

7 January 1909: the plate that showed bent lines

John Evershed arrived at Kodaikanal as an assistant to the director. His Royal Astronomical Society obituary, written by F. J. M. Stratton, says he was appointed in 1906 on the initiative of Gilbert Walker, then Director of Observatories in India. IIA’s D. C. V. Mallik, writing in the journal Resonance, gives 1907 as the year he came to Kodaikanal and 1911 as the year he became director.

Evershed’s paper, “Radial Movement in Sun-spots”, appeared in the Monthly Notices of the Royal Astronomical Society in 1909, dated from Kodaikanal on 6 February. He used a Rowland plane grating, a ruled glass plate that spreads sunlight into a spectrum, in its fourth and fifth orders. The dark lines in a spectrum are called Fraunhofer lines, and a gas moving toward or away from the observer shifts them slightly in wavelength.

Where the spectrograph slit crossed a sunspot, the lines were slightly bent compared with the lines from the neighbouring solar surface. Evershed wrote that the first plate showing this clearly was exposed on 7 January 1909. On it, the lines on each side of the spot’s penumbra (the lighter, fibrous ring around the dark core) were shifted by 0.027 angstrom, a unit of one ten-billionth of a metre. Evershed converted that to gas receding at 0.85 kilometres per second on one side and approaching at the same speed on the other.

From 150 spectra to a radial flow

Evershed’s first reading was that the gas in the penumbra rotated as a whole around the spot’s centre. He then tested it. In the paper he reports about 150 spectra taken since 7 January, covering seven spots in the northern solar hemisphere and four in the southern. The shifts vanished when a spot was within 10 degrees of the centre of the Sun’s disc and were strongest between 30 and 50 degrees from the central meridian. No shift appeared when the slit cut across the spot at right angles to the line toward the disc centre.

Those results fitted a flow moving outward from the spot’s centre and parallel to the solar surface, and ruled out a vortex. Evershed also used the turning of the solar image in the heliostat (the mirror that tracks the Sun) through the day to find the line along which the shift disappears. His table of seven determinations averages 90 degrees from the direction of the disc centre, as an outward flow requires.

His largest measured shift was 0.040 angstrom for a spot 40 degrees west of the central meridian, which he converted to a motion of 1.9 kilometres per second along the Sun’s surface at the outer edge of the penumbra, about 32,000 kilometres from the spot’s centre. He noted a conflict with George Ellery Hale’s discovery of the Zeeman effect (the splitting of spectral lines by a magnetic field) in sunspots, which seemed to demand circular motion. He listed an unexplained difficulty too: the flow seemed to accelerate outward and then stop abruptly at the penumbra’s edge.

Stratton’s obituary adds that Evershed was elected to the Royal Society in 1915 and received the Royal Astronomical Society’s Gold Medal in 1918, largely for this work. It also describes the motion as accelerating outward in the reversing layer, where the dark lines form, and inward in the higher chromosphere.

What the Evershed effect looks like now

Mallik calls it the earliest successful observation of velocity fields from a complex magnetohydrodynamic process, meaning the behaviour of electrically conducting gas threaded by magnetic fields. IIA’s museum page goes further and calls it the first astrophysical observation of an interaction between plasma and magnetic field, a description that is the institute’s own.

A 2009 summary by K. Ichimoto and the Solar Optical Telescope team on Japan’s Hinode satellite says the telescope resolved the flow into narrow channels with nearly horizontal magnetic fields. It reports a flow speed close to the sound speed at the Sun’s surface, with hot gas rising at the inner end of each channel and sinking at the outer end. The same paper says there was still no consensus on the physical origin of the fine structure of the penumbra.

A century of plates, scanned

The plates and films are the other half of the record. IIA says the historical photographs have been digitised and are available through a public portal built at the Inter-University Centre for Astronomy and Astrophysics (IUCAA). The institute’s press release of 2 April 2024 puts the repository at 1.2 lakh (120,000) digitised solar images.

The Ca II K plates have been scanned more than once. Chatzistergos and colleagues describe three digitisations, of which only the last two covered the whole solar disc. An earlier one, by Makarov and colleagues in 2004, covered 22,158 photographs from 1907 to 1999 as 8-bit images. The later one, by Priyal and colleagues in 2014, covered 48,928 photographs from 1904 to 2007 as 16-bit images, taken with a 4096 by 4096 pixel camera. That start year is three years earlier than the 1907 on IIA’s data page, and the sources do not explain the difference. Priyal’s team says the scanning was done in a room with controlled temperature and humidity. The 2019 paper concludes that the 16-bit scans improve on the earlier ones.

Different groups analysing the new scans have reached results that differ from each other, and Chatzistergos and colleagues found timing errors in the observation times in both digitisations. Some of the problems appear in both digitisations, which the authors take to mean they come from the plates themselves. A 2025 IIA research highlight describes a reprocessing of Ca II K images from 1907 to 1996, covering Solar Cycles 14 to 22, in which Rieger-type periods, short repeating swings in plage and network area, were significant in every cycle and in both hemispheres.

The observatory also kept hand-drawn charts. A 2020 paper by B. Ravindra and colleagues says sunspot observations began in 1905 and that the observatory holds 111 years of drawings across ten solar cycles, bound in volumes of six months each. Their counts run about 40 per cent below the Royal Greenwich Observatory’s, so they normalised the Kodaikanal numbers to the Greenwich series.

What the observatory does today

The record continues with newer instruments. IIA lists a hydrogen-alpha telescope installed in October 2014, a Twin Telescope that imaged the Sun in Ca II K and white light from February 2008 to 2012, and the WARM telescope, which has made full-disc images since January 2016 and has carried a Ca II K filter since February 2017. A solar tunnel telescope, with a 60-metre underground tunnel, serves spectroscopy.

Radio work began with a 100 MHz interferometer in 1952, which IIA calls the earliest solar radio observation in India. An ionosonde, which sounds the upper atmosphere with radio pulses, was installed in 1955. M. K. Vainu Bappu became director in 1960, and IIA’s history page says that when he established the IIA in Bangalore in 1977, most Kodaikanal astronomers moved there.

The ground record now has a space companion in ISRO’s Aditya-L1, and IIA also runs the high-altitude observatories at Hanle in Ladakh.

Visiting

IIA’s pages, checked on 28 October 2026, put the observatory 4 km from Kodaikanal town. The nearest airport is Madurai, 130 km away, and the nearest railway station is Kodaikanal Road, 80 km away. The museum page says the astronomy museum, 400 metres from the main gate, is open daily including government holidays from 9:30 to 12:00 and 14:00 to 16:30, with a night skywatch from 6:30 to 7:30 pm. The visiting page, last updated in November 2024, asks groups to submit a form at least 15 days ahead, gives 9:00 to 13:00 as Sunday hours and allows up to 90 minutes on campus. The two pages give different hours, so check with the observatory before travelling.

The stamp, released on 16 May 2025 at the IIA in Bengaluru, also carries a solar image taken at Kodaikanal on 6 May 2024.