How do we know what stars and other celestial objects are made of? Astronomy, spectroscopy and Sydney Observatory

Multi-prism spectroscope made by Hilger

Government Astronomer Henry Chamberlain Russell purchased this spectroscope for Sydney Observatory while in London in 1875. It was made by the then new firm set up by instrument maker Adam Hilger. Powerhouse Museum collection H9974

In a recent post on the Powerhouse Museum’s collection blog my colleague Debbie Rudder explains how chemists could discover new elements by examining the spectrum they emitted when heated. Each substance emits its own characteristic ‘barcode’ of lines at different colours or wavelengths. These can be studied by a device called a spectroscope that uses prisms (or gratings) to spread light into its constituent colours.

Astronomers quickly seized on this new technique to examine distant stars and nebulae as it provided for the first time a way of discovering the constituents of these unreachable objects. Strangely, it was not the major professional observatories such as the Royal Greenwich Observatory that took the lead, but wealthy amateur astronomers. Among the most prominent of these was William Huggins, later Sir William, who had a private observatory at Tulse Hill in Lambeth, South London.

Huggins published his first paper on spectral lines in 1863 and by the next year made the important discovery that some nebulae were made up only of gas and did not contain any stars. Soon he was measuring the velocities of stars using the Doppler Effect, the slight shift in the wavelengths of spectral lines due to velocity. He also experimented with photography to record the spectra that he was observing.

NSW Government Astronomer Henry Chamberlain Russell was clearly aware of developments in this new field of spectroscopy for while in England in 1875 he purchased a spectroscope. The spectroscope shown above was made by the then newly set up firm of Adam Hilger. Russell used the spectroscope on at least one occasion when he took it together with one of the large telescopes from the Observatory up to the Blue Mountains. However, following the lead of Greenwich, Russell kept on with routine astronomical tasks and did not take up the challenge and opportunity offered by the spectroscope.

Hilger 1915 spectroscope detail

The Hilger spectrograph purchased by Sydney Observatory in about 1915. Powerhouse Museum H9955

In the early 1900s Sydney Observatory purchased another spectroscope from the same firm. This spectroscope, H9955 in the Powerhouse Museum collection, was made in 1913 or soon afterwards and shows considerable advances in design in that it is clearly designed to be connected to a telescope and in that it is designed not just to allow visual views, but also to record spectra photographically. Hence, technically it should be called a spectrograph and not a spectroscope. A cross-sectional blueprint that came with the instrument shows that it contains two prisms to spread the light coming from the telescope. As shown by the picture below, Sydney Observatory astronomers did use the spectrograph on appropriate occasions.

Harley Wood using the 1915 Hilger spectroscope

Government Astronomer Harley Wood using the later Hilger spectrograph, probably in the early 1940s, on the 29-cm telescope in the Observatory’s south dome. The eyepiece he is using allowed him to examine light reflected off the face off the first prism and hence to check if the star, comet or other object is properly positioned to shine light into the spectrograph. Picture Powerhouse Museum

In Australia spectroscopy and astrophysics only began to be seriously pursued with the establishment of Mount Stromlo Observatory, near Canberra, in the mid-1920s. The Powerhouse Museum has a number of important spectrographs from Mount Stromlo Observatory, but that is another story or another blog post. Here we just note that to today’s astronomers have a variety of techniques at their disposal, but spectroscopy is by far the most important.

Russell’s 1875 Hilger spectroscope is most likely the earliest spectroscope purchased for astronomy in the country and as such it is of crucial importance and of great significance in the history of Australian astronomy.

11 responses to “How do we know what stars and other celestial objects are made of? Astronomy, spectroscopy and Sydney Observatory

  • I’ve just checked and ‘The spectroscope and its applications’ by J. N. Lockyer (1873) is in the Powerhouse Museum library.

    • Great news.
      What you should also looking for is an important document made after May 1877 and by the end of 1878. This contains information and a drawing of the solar spectrum that was sent by “Dr Higgins” to Russell. This is quite different, apparently, than than the famous one published by Higgins 1868. (See “Notes of an Astronomical Experiment made on the Blue Mountains, near Sydney, N.S.W., Obs.,2, 370 (1879) [RAS Journal] pg.373.)

      The letter and drawing is probably far more valuable than the spectroscope itself!

      Regards, Andrew

  • Thanks for the comments Andrew, on both Russell and Janssen. Perhaps the book “The Spectroscope and Its Applications.”, by J.N.Lockyer (1873), you refer to is in our library? On return from London I will look for it. I am also wondering if this is the object shortly to be put on display at the Observatory.

    • Hi Tona

      I’ve never seen this 117 page book in your library. Like most, it seems to be a bit of a lottery in what still lies in the collection of Syd Obs books. It was however a standard text of the day, and gives a perspective on the level of knowledge.

      You can also download this whole 2nd edition book from the “Open Library” (as pdf or ePub) at the webpage at;


      (Even pg.126 page has an advertisement about Howard Grubb and the Great Melbourne telescope.)

  • Ah. (It is amazing what you can remember with a bit of contemplation.)

    Calibration of spectroscopes were often made by an instrument known as a cathetometer, which is a horizontal measurement device with a telescope (or microscope) that could be raised or lowered. The eyepiece had a calibration scale, which could be read to considerable precision. In the 1870 and 1880 it was used to calibrate thermometer and barometers (manometers), by measuring the height of the meniscus. From measuring two known points, any instrument could be quickly calibrated. Such a device was also useful in spectroscopy calibration. Usually hot a flame had some heat element or compound, whose emission line of known wavelength was seen through the spectroscope. I.e. salts of potassium, calcium or sodium. (You could also calibrate on a solar spectrum.)

    Russell must have used something like this to calibrate for his meteorological work. I do not know if Sydney observatory id this on this Hilger device or had a catetometer or something similar. Melbourne observatory certainly had one in 1884 made by T&S — and possibly one earlier from the Lockyer eclipse expedition in 1870.

    Now, I also should have mentioned Norman Lockyer, who discovered helium in the solar spectrum. Lockyer was one of the first to become interested in stars and “electromagnetic spectroscopy” in the 1860’s, and firstly applied it to Fraunhofer spectral lines found on the sun almost sixty years earlier than him (1802, actually). In the sun spectra, he could not account for the so-called yellow d3 line at 587.6nm., which Lockyer deduced was a new element that he named helium.
    This in turn lead him and others to venture to total solar eclipses. The eclipse it noted before was in India in 12th December 1871, and this is where the Melbourne spectroscopes we believed come from. Observation revealed the He line and the chemical nature of the chromosphere via the flash spectra. (Tona might be interested that this yellow line also was made by Janssen at the same eclipse, who was mention in the earlier Chamonix article in France.)

    The ‘tell-all’ book was “The Spectroscope and Its Applications.”, by J.N.Lockyer (1873), which Russell and Ellery became aware of, and decided to embark on possible southern observations. Ellery’s culmination was with the emission spectra of Gamma Velorum in 1879.

    Remarkably, the failed 1869 spectroscope for doing solar work predates all of these, and suggests some collaboration or verification of Locker’s work.

    Verification of some of these works can be found in Barry Clarke’s ASv article; “Melbourne Observatory : New Inventories of Associated Items, Status of the Site. and Proposals for Heritage Conservation and Development” (2006) [See pg. 108-109, 143, especially.] [This mentions the pre-Stromlo inventory too.]


    • I found an image of the cathetometer mentioned above in the PowerHouse Curatorial Research Series “Early astronomical telescopes and equipment at Sydney Observatory.” (See attachment image.)

      In “Notes of an Astronomical Experiment made on the Blue Mountains, near Sydney, N.S.W.“. By H.C. Russell written on 26th July, 1877 ; Obs., 2, 370 (1879) describes;

      “Of the quality of the 7¼-in. telescope I need say nothing more now than that it is a first-class instrument, as these results will show, and well adapted for the purpose. The spectroscope was made by Hilger, of London, and at the time (1876) it was considered the most powerful and perfect one in the world.

      At 11.30 A.M.. of the 17th October I began the spectroscopic work, and was at once struck by the clear definition of the lines : the gain in clearness over the Sydney atmosphere was very striking; but the next moment I asked myself, where were the lines? instead of seven lines between the two D’s there was but one, and on either side of them, where it is usual to see a host of lines, the spectrum was remarkable for their absence. The line D1, however, which I discovered in 1877, was visible definitely, but very fine; and on very carefully examining the space between the D’s there was a shading such as would be produced by very line lines, too faint to be seen separately. These observations were made about noon, and I did not again look at the spectrum until near sunset, when a great change had come over it. Six lines were visible between the D’s and the usual host on either side; they were more clearly defined than I had ever seen them before, and as the sun approached the horizon they thickened very rapidly…

      The 18th October proved a cloudy day and night, with a light easterly wind or sea-breeze. Towards the morning of the 19th, the clouds began to break, and soon after 9 A.M. the sun shone out : at 10.30 A.M. there were a few passing clouds only; but the sky was very white, as it is in Sydney, and looking at the spectrum I found six lines between the D’s. The cloud were disappearing fast, and by 11 A.M. it was difficult to make out the six lines; they were fading fast; on either side there seemed to be a shading such as was seen on the 17th. At noon the air-lines had faded, and the spectrum appeared as in figure 1 (that is, without air-line those shown being evidently part of the so ar spectrum). At 4 P.M. seven lines could be distinctly seen between the D’s, and the first set of measures were taken. At 5.20 P.M., when the air-lines were getting better defined, so lines were measured between Ca and D₁, where only three had been seen at noon; their positions are shown in figure 3 as well as the D lines seen at the same time. After these measures had been taken, the micrometer was removed and an eyepiece that defined a little better was used; and now 24 lines could be counted between Ca and D1, and for the first time an eighth line was seen between the D’s ; and just as the sun was setting, the air remaining perfectly steady, 26 lines were counted where only 20 had been measured ; in fact it seemed as if new lines were coming into the spectrum every minute, and one ceased to wonder at the loss of light, when the lines, like so many strokes from a pen, came rapidly one after the other, each one definitely blotting out a part of the sunlight.

      Another fact was now clearly made out: in Sydney, owing to the unsteady air when the sun is near the horizon, it had been supposed that the thickening at D1 was all in that line, and that D2 was lost; but now it was evident that it was D3, which became thick and equal to D3, so that only the faintest line of light separated the two.

      He concludes;”Knowing the proverbially clear atmosphere of our mountains I was anxious to see if there was still less there than in Sydney; and on a fine clear day, with dry wind at noon, all the air-lines in the part examined seemed wanting; but as the Sun descended to the west the lines gradually appeared, and when he was near setting, and the vapours therefore began to condense about the hills, the lines were seen conspicuously ; or if an easterly wind brought the sea air on to the mountains the lines could be seen at noon. It would seem therefore that the cause of these lines is water vapour; and if any of the 24 lines seen in London are due to gases which are the product of manufacture, we in Sydney have so far not contaminated the air in this way, for all that could be seen in Sydney are also visible on the mountains.

      The definition of the spectrum lines on the mountains was a sight never to be forgotten; and it is difficult to assign a value for the gain, but it cannot be less than 50 per cent.; the lines seemed so hard and definite that there was no difficulty in measuring or in dividing close lines that under other circumstances could not have been done.”

      In my searches over three weeks or so, these seem the only observations I could find as even remotely relevant.

      Andrew James. 

      • It is also interesting that the chemical elements of these lines were known. I’d assume the calibration with the cathetometer would have been made using a Calcium compound (probably Calcium carbonate) or Sodium or Potassium Chloride. (Sodium Chloride being just household salt.)

  • Now this is an interesting story.

    Firstly off the top, there is little of any significant observations when using these spectroscopes. I know of no significant or useful spectra made at Sydney observatory of the brightest stars or other celestial objects during the time of Russell. (I’ve always assumed it was mostly a showpiece for the prestige of Sydney Observatory and H.C. Russell’s wide-based scientific programme.

    I think the main reason for the purchase was almost solely due the remarkable star Eta Carinae in the 1840s to 1861. By 1875, the star had diminish in brightness as such, where any spectra was not very useful in terms of resolution. (Much of the Eta Car spectral work later came from Madras Observatory in India by Captain W.S. Jacob (and others) who also had a fascination with astronomical equipment.

    H.C. Russell made many magnitudes estimates of eta Car, (one of 5th May 1871 comes to mind I.e. “Observations on the stars and nebula about η Argus.” (1871). Another Trans. Roy.Soc. NSW, 5, 15 (1871)), and there is no doubt he wanted to take advantage of the southern location to do more useful work of this star and the general region surrounding it. This, I’d think, would have been the motive to getting the Hilger spectroscope. (This spectroscope would have been of interest to chemists, and could be used to examine by flame spectroscopy to identify basic elements or compounds. Some work might have been achieved by this method, and no doubt, might have been used for calibrations. Another calibration method was the solar spectra. I’m unsure if either was ever done with this instrument. Some details were discussed on this of short-lived Royal Society of NSW; Section B, which discussed scientific advancements or experimentation. [Sadly the written minutes of this are now missing, so we have no way of knowing the level of the discussions here ] )

    It is also interesting that he became immediately interested in the colour of Eta Car. I.e. 31st January 1876 he discusses the variations in the redness of the star (then at about 6th or 7th magnitude.) It seems Russell may have made or attempted some spectral observations during 1887 and 1888, which according to David Frew were never published and were seemingly “lost” to us. Most of the spectral work suffered mostly due to a lack of aperture, especially due to the crudeness of the optics; especially mirrored surfaces like used with prisms.

    Perhaps the best and earliest documented article on spectra from Australia was obtained by Robert Ellery on 14 January 1879 of the star Gamma Argus (now Gamma Velorum). He found on two evenings three lines; two yellow at 576.0nm and 564.8nm, and one blue at 468.2 (HeI line). This was obtained by the Grubb spectroscope using the aperture of the Melbourne reflector. (“The spectrum of [gamma] Argus”, MNRAS., 2, 418) Ellery also used during a direct vision spectroscope around 1889 and published in; “Preliminary spectroscopic survey of southern stars, made at the Melbourne Observatory with a Maclean direct vision spectroscope.” MNRAS., 49, 439-445, Supplement (1889). By the end of the century, Melbourne Observatory also had a solar spectroscope, that apparently did not work.

    However, the most important southern spectroscopic observations were mostly made in South America during the 1890s, at the American university “outstations” like at Aquipea. Some spectral work was done between 1900-1920 at the Cape Observatory in South Africa, when Eta Car had “bottomed-out” at 8th magnitude or so. The Cape observed no spectroscopic events during this time for Eta Car.

    According to Roberta Humpherys at the University of Minnesota “Eta Carinae’s Historical Spectra : Complications for the Binary Model,”;

    “The first photographic spectrum was obtained as part of Harvard Observatory’s objective prism survey during eta Car’s second eruption in 1892-93. It is described by Cannon (1901),”

    After the main maxima, the lines that were the most prominent in eta Car were mostly from He I, [N II], [Ne III], [Ar III], [Fe III], and observations of changes in these lines are well documented since about the 1900s. We know little of these spectral characteristics during the time of maxima when the star shone on or above 1st magnitude. After this photographic records of spectra only start in the 1940s onwards.

    As T.J.J. See (MNRAS., 57, 541 (1897), right describes the problem of obtaining spectra, saying; “It will be readily be understood that the atmospheric spectra of stars seen at our station [Lowell Observatory], 7,600 ft [1.5 km.] above sea level, are much less troublesome than at at places where the atmosphere is denser… , and hence in this work we have been careful to select nights of good seeing.” This same issue applied to Russell, hence his observations in the Blue Mountains to gain altitude.

    Thanks for this article. You’ve inspired me to do some extra work in this particular field. [I think you will find more detail here in the early chemistry work at Sydney University. There was a sideline interested in chemical analysis at the time. How these instruments were calibrated is of particular interest to me.]


    Note 1: The oldest astronomical spectroscope was made by Browning or Hilger, and had seven glass prisms. This was obtained by Melbourne Observatory in 1869. Others were obtained there in the early 1880s. I knew of these spectroscopes from a visit in Devon, UK in 1998, and they were apparently ‘donated’ by Sir Norman Locker, as Ellery says “…for the use in your hemisphere.” (This was after a British eclipse in the expedition in the region in 1871 or 1872.) [Sorry. This sadly discounts the claim of; “Russell’s 1875 Hilger spectroscope is most likely the earliest spectroscope purchased for astronomy in the country and as such it is of crucial importance and of great significance in the history of Australian astronomy.”]

    Note 2: One of the documents, which I’ve only seen and glanced at once, is H.C. Russell’s “Aurora australis.” (1897) Record. Ref. 209/185 I think this too has a short discussion on spectra.

    • > Do we know what happened to the “seven prism” presumably Browning at that date, spectroscope?
      Was this the one used by Le Sueur for his observations of eta car from Melbourne?

      • Hello Ken. Thanks for your interesting questions to which I am unsure of the answers. I will try to find out and post a proper reply.

        • Actually there may be some confusion around this early spectroscope…
          p126 advert confirm that Grubb actually supplied a spectroscope to work with the GMT….maybe this it the one being discussed??

Leave a Reply

Your email address will not be published.