Artificem egregium, speculatorem rerum, coelestium callidum.
Such were the words bestowed upon Alvan Clark snr. in 1874 by Harvard College Observatory, in recognition of his impressive advances in telescope optics. It was a fitting accolade for a man, who, despite very humble origins, ascended to international greatness. Empowered with the trained eye of a portrait painter, Clark went on to become a gifted telescope maker and, in time, entrusted his sons to the task of building the largest refracting telescopes the world had ever seen.
The Clark refractors were used to make some of the most famous astronomical discoveries of the 19th century. On a freezing late January night in 1862, the Clarks were performing routine tests on a newly completed 18.5 object glass. They were trying to gauge how much off-axis glare the instrument exhibited by timing how long the light from Sirius was perceptible before the star was in view. While the bright star was still behind a corner, Alvan Graham Clark noticed an eighth magnitude ‘spark’ appear a full three seconds before Sirius came into view. This was the first recorded observer of Sirius ‘pup’ and a testament to the quality of optics used to divine its secret.
In 1877, Asaph Hall used the 26-inch Clark refractor at the US Naval Observatory to discover the two Martian satellites, Phobos and Deimos. Then there’s Amalthea, a tiny satellite of Jupiter, measuring just 250 kilometres at its widest extent, which was detected visually by E.E. Barnard using the newly installed 36-inch Lick refractor atop Mount Hamilton, on the faithful night of September 9, 1892. The Clark refractors were used fruitfully to greatly extend our knowledge of double stars. Prominent among them was Sheldon Wesley Burnham, who, after acquiring a 6-inch Clark of the finest specification, uncovered 451 new pairs from 1872 to 1877. Burnham’s extraordinary success with such a modest instrument embarrassed the astronomical cognoscenti, who had mistakenly believed that essentially all the binary stars visible to the instruments of the day had been discovered. Using the great 40-inch Clark refractor at Yerkes Observatory, Wisconsin, Robert Grant Aitken was able to measure some astonishingly close binary stars, some with separations as diminutive as 50 milli-arcseconds!
And in Britain, prominent Victorian amateur astronomers, such as Sir William Huggins and the Reverend William Rutter Dawes (the eagle eyed), used medium aperture Clark refractors to advance the cause of spectroscopy and planetary science, respectively.
The optical and mechanical quality of the Clark refractors guaranteed their success in the United States and abroad. But how good were the Clark lenses? Indeed, more broadly, the curious individual may legitimately enquire as to the differences between the quality of 19th century glass and their modern counterparts. The correspondences I have had with a few owners of 4-inch Clark refractors invariably attest to their very high quality. Daniel Schechter, a physician based in California, is an avid collector of antique Clark refractors, including a wooden-tubed 4-inch dating from 1860 featured above. “I had this Clark objective expertly tuned and tested,” he said, “and it turned out to be better than 1/8 wave after expert spacing and collimation.”
Larger lenses, of course, get exponentially more difficult to figure accurately and, as we’ve seen, the Clarks made many large telescopes. Owner testimonies, while informative, can only tell half the story though. What is required is an objective testing procedure that can quantify how good these objectives really were. To that end, I contacted Dick Parker, who has ground and built his own telescopes for decades and now runs a telescope making workshop at his home in Tolland, Connecticut, USA.
Proof of the pudding
Dick recounted details of work he carried out on the testing of a 5 inch f/15 Clark objective. “In September, 2010 I had an opportunity to test a 5 inch diameter achromat lens,” he said, “which is currently owned by an antique telescope collector. It is dated to c.1915 and appears not to have ever been installed in a telescope.”
Dick made use of the so-called an auto-collimation test, which yields a null return for an optic bringing light originating at infinity to a single focus point. Collimated light, in optical terms, means light rays that are parallel. The object glass makes its own collimated beam to simulate light coming from stars at infinite distance. Curiously, this test was actually adopted by Alvan Clark & Sons and they reported this testing technique in Scientific American Supplement No. 932 for November 11, 1893.
Dick took some time to explain the procedure. “For a practical description of the test, a pin hole source of light (artificial star) is placed at the focus of the lens. The light travels to the lens in a diverging cone, then passes through it, where it is refracted to a parallel bundle of rays (collimated). The collimated light then is intercepted by an optically flat mirror and reflected back through the lens, where it is refocused back near the original source. At this focus location, the returning focused beam can be examined by an eyepiece, knife edge, Ronchi screen or other suitable testing device.
“I chose to use a Ronchi screen,” he said, “which consists of a series of consecutive opaque and transparent lines, etched in at a very fine spatial frequency. I adopted a screen with 133 lines per inch, which was placed just behind (further from the lens) where the returning beam comes to focus and examined. Then the screen was placed just before the beam comes to focus and re-examined. If the rays from the lens are brought to a perfect focus, what will be seen is a series of lines across the lens that will be straight, parallel, and equally spaced. Any curvature of the lines will reveal a defect in the lens. The views with the screen just behind the focus (further from the lens than the focus) should be the same as views with the screen in front of where the rays come to focus. The light source used had a wavelength 565 nm (green).
Figure 1 and Figure 2 shows the test return with the Ronchi screen just behind and just in front of where the rays come to focus, respectively. What should be visible by comparing the two is slight bowing of the lines. Notice that the lines appear to bow inwards toward the outer part of the lens in figure 1 and outward toward the center of the lens in figure 2.
This means that light coming through the center of the lens focuses shorter than light coming from the outer parts of the lens. This is called spherical aberration.
What is also noticeable in Figure 1 is increased exaggeration of this bowing toward the center third of the lens. This is an indication of a local condition where the centre of the lens focuses just a bit shorter than it would from spherical aberration alone.
Figure 3 was made with the Ronchi screen at focus so that one bar of the screen acts as a single knife edge. This makes for a very sensitive test for local conditions. If all the rays were focusing at a common point, the surface would appear uniformly grey. Notice the centre appears as if there were a depression in it. This is indication of a short focusing center.
What does this tell us about the lens? It does mean that the lens is not “perfect”, however the lens should provide very fine images were it to be installed in a telescope. By comparing to a known optic with optical path difference between the centre and edge due to spherical aberration equal to ¼ of the wave length of light, tested the same way by auto-collimation, I would judge this lens to have spherical aberration just about ¼ wave or better. “
So, according to Dick Parker’s optical bench tests, this particular Clark lens met or exceeded the minimum quality contemporary opticians would deem useful. That’s quite an acceptable figure, given that the company was probably working to the minimum standards arrived at by Lord Rayleigh in 1879. Indeed A. E. Conrady’s later book “Applied Optics and Optical Design” which first published in 1929,arrives at much the same conclusion.
Of course these telescopes also had other advantages over their modern short focus counterparts, with low dispersion glass. Dr. Juergen Schmoll, an instrument scientist based at Netpark, Durham, kindly informed me that the Coefficients of Thermal Expansion (CTE) of both traditional crown and flint glass are roughly half that of any modern ED glass, and nearly one third that of fluorite! That’s a very significant revelation, as elements with lower CTE values distort less during cooling and thus hold their figure better during acclimation. In addition, a lens that morphs as it cools will be more difficult to focus accurately as it will introduce aberrations similar to spherical aberration into the optical train. This is yet another factor contributing to the greater stability of the achromatic images compared to their modern counterparts and goes some way to explaining why they delighted so many observers over decades and centuries. Long live the Clarks!
Warner, D.J., Alvan Clark & Sons, Artists in Optics, (1968), Smithsonian Institution Press.