The Historical Growth of Telescope Aperture

A simple exponential growth least‐squares fit to all historical data in Table 1 yields an aperture doubling time, t_2×, of 52 yr and a 1900 diameter D₁₉₀₀ = 1.5 m, with a logarithmic dispersion of 0.22 (±67% in D). These values are slightly inflated by telescopes denoted "s" in Table 1 and shown as crosses in Figure 1. Excluding them from the fit reduces the dispersion to ±50% and yields t_{2 ×} = 50 yr and D₁₉₀₀ = 1.4 m. These giant but barely functional instruments tended to have twice the aperture of their more productive large contemporaries, but, as history records, delivered such inferior images or were so difficult to operate that their returns were limited. Their sizes exceeded the limits allowed by their concepts or by the technology of their times. Notable examples are some of Hevelius and Cassini's 150 foot long "lunettes à faire peur aux gens," which so frightened Chrysale in Les Femmes Savants (Molière 1672, II: 7); Herschel's 1.2 m "40 foot reflector;" and the Earl of Rosse's 1.8 m "Leviathan of Parsonstown."³ The disappearance of these valiant but premature efforts after the mid‐1800s is no doubt a consequence of experience and of prudence instilled by their forbidding cost. The data for these telescopes are disregarded in what follows.

The first optical instruments were refractors: nature produces transparent optics—eyes—but no mirrors. Aperture size for the early, simple‐lens refractors grew with t_{2 ×} = 25 yr during the 17th century. Had this trend continued, we would have had 100 m telescopes by 1900! The technology of early refractors allowed relatively rapid growth. But this growth was thwarted by an inherent limitation: the very large f‐ratios needed to limit chromatic aberration, typically f/150, eventually imposed such length requirements on the instruments that they became unmanageable. Important discoveries in planetary astronomy were, nevertheless, made with these instruments, which prompted King to comment that "the success of the long telescopes of the seventeenth century was due, very largely, to the painstaking and persistent efforts of men like Hevelius, Huygens and J. D. Cassini. Indeed, after Cassini's death in 1712, his successors were unable to see what he had already discovered, let alone add to the list, and the telescopes gradually fell into disuse" (1955, p. 133). New technology was needed.

New technology had already emerged in 1668 with Newton's reflecting telescope (Newton 1672, 1718; Birch 1756), which came to refractors with the invention of the achromatic objective (Dollond 1758). These two types of telescopes were to compete for preeminence for 2 centuries. While early reflectors, at ∼10 cm, coincidentally matched the aperture of the largest functional refractors of their time, achromats only reached that size a century later, for well‐known reasons: Newton's anathema for refractive optics and the technical difficulty of producing large‐flint blanks.

The switch to the inherently more promising achromat technology at first forced a ∼4 factor drop in refractor aperture and led to slower aperture growth than allowed by simple‐lens telescopes. This shows that changes in technology can affect both the zero point and the slope of the D(t) relation, at times in directions that do not immediately appear advantageous for astronomy. Yet historical records show that the smaller and slower growing achromats surpassed their larger single‐lens ancestors in performance, in particular in image quality and ease of operation. Once mastered, achromats fostered a brilliant series of increasingly large telescopes, which culminated with the Lick and Yerkes refractors before also being stopped by inherent limitations. The need to limit residual aberrations again led to overly long tubes, which were difficult to support, house, and drive; and much larger glasses of sufficient quality were impossible to produce. A last hurrah was hailed with the 1.2 m objective of the 1900 Paris Exhibition, which was stationary and fed by a steering flat, and through which visitors were invited to "voir la lune à 1 mètre." It served astronomy little. Aperture growth henceforth had to rely on reflecting telescopes.

The growth of large refractors during the 18th and 19th centuries is well represented by a simple exponential, shown by a dashed line in Figure 1, with t_{2 ×} = 45 yr, D₁₉₀₀ = 0.9 m, and a dispersion of ±18%.

The 58 historical reflectors in Figure 1 (filled circles) define a curve of growth with t_{2 ×} = 50 yr, D₁₉₀₀ = 1.6 m, and dispersion of ±37%. The doubling time is essentially the same as that for refractors, but the mean diameter is larger by a factor of almost 2. Aficionados of historical astronomy will recall the heated competition between reflectors and refractors throughout the 19th century, when advocates for the latter claimed that they surpassed in performance reflectors of twice their aperture. Thus, in his 1874 deed of trust, which would give birth to the 36 inch (0.91 m) refractor, James Lick intended it to be "...superior to and more powerful than any telescope yet made" (Shapley & Howarth 1929; Osterbrock, Gustafson, & Unruh 1988), despite the fact that larger reflectors had been around for a century. He may have been right.

The dispersion of the residuals about the exponential fit for historical reflectors is large (±37%). A possible explanation is ventured in § 3.4.

Beginning with James Short and William Herschel in the second half of the 1700s, some telescopes—all reflectors of course—were built with the explicit goal of being "the biggest in the world," and no efforts were spared to make them so. "I can now say that I have the best telescopes that were ever made," Herschel said in 1782, after the completion of his 20 foot telescope (Lubbock 1933, p. 116). Thhe same sentiments must have been felt at the first lights of the Lassell, Crossley, Hooker, Hale, Keck, and other "biggest in the world" telescopes. That breed of "frontier" instruments fires the imagination, drives technology, and tends to dominate astronomy.

Other, second‐tier merely "large" reflectors were built to provide competitive capabilities within constrained budgets, and with the benefit of advances made by the first group. Telescopes of this breed tended to arrive in bursts at ∼35 yr intervals in the 20th century, at times in commercial series, such as the Warner and Swasey and the Grubb‐Parsons telescopes of the 1920s and 1930s, or through collegial, and hence mutually supportive, efforts such as the 4 m family of the 1960–1970s and the 8 m's at the dawn of the 21st century.⁴

It appears that aperture growth followed different laws for these two breeds of telescopes. Frontier reflectors naturally led large ones in size by a few decades. Figure 1 also suggests—especially with the eye‐drawing line shown through the data points—that the largest functional reflectors, from instruments by Short and Herschel to the Keck telescopes, define a sharp, distinct, and strictly exponential upper limit to the D(t) distribution. The smaller 20th century telescopes of Table 2, plotted as filled circles in Figure 1, help to show the distinctness of the frontier telescopes, whose exponential growth is characterized by D₁₉₀₀ = 2.3 m and t_{2 ×} = 48 yr, with a dispersion of only 14%. The relation

must then describe how fast the advances in technology have allowed these telescopes to grow.

William Herschel's heroic efforts at building increasingly large reflectors illustrate well the onset of technological limits. Six of his major instruments appear in Table 1. With his 0.35 m aperture, 20 foot (6 m) telescope in 1778, Herschel surpassed Short's 28 yr old record. His attempts at a 0.6 m mirror for a 30 foot (9 m) telescope failed catastrophically in 1782—the speculum‐melting furnace exploded—and he redirected his work to a more modest 0.45 m "large 20 foot" completed in 1785, which was very well used afterwards. Herschel's 1.2 m "40 foot" (12 m) was an extraordinary feat. But, by his own admission, the 20 foot (6 m) was scientifically preferable because it delivered better images and was far less cumbersome to maintain and operate. "To look through a telescope larger than required is a loss of time which an astronomer has not to spare" (Herschel 1798). And King comments: "The paucity and irregularity of observations with the 40 foot leave no doubt that the great telescope failed to meet its maker's expectations." By the 1790s, technology (mirrors, mountings, etc.) was simply insufficient to allow the greatest telescope maker to profitability go beyond a 0.5 m aperture. For the same reason, Hale abandoned his project for a 25 foot (7.6 m) telescope (Hale 1928)—a model is still at the Carnegie Observatories in Pasadena—and decided on a 200 inch (5.08 m) telescope, the construction of which was delayed by World War II.

If the dichotomy between "frontier" and "large" telescopes is accepted, large reflectors show an interesting trend in Figure 1. There can be little doubt that since the 1950s they have grown faster in aperture than their frontier cousins. The curve drawn through their data points represents a growth for which t_{2 ×} = 47 yr until the end of World War II and then decreases exponentially with time, e‐folding in 70 yr, to yield t_{2 ×} = 20 yr in 2000. Whether or not this acceleration is real and deserves further analysis, this last value is the lower limit allowed by the data for the current aperture doubling time.

Table 1 and Figure 1 include five future extremely large telescopes (ELTs), ranging in size between 20 and 100 m, and enthusiastically promoted in the US, Europe, and Canada. Others may soon appear. Their necessarily tentative schedules define an aperture growth for the ELT projects characterized by t_{2 ×}∼5 yr if D₂₀₀₅ = 12 m (e.g., Large Binocular Telescopes [LBTs]) is imposed, with a dispersion of factor of 2 in D. This line is dashed through the ELT points in Figure 1.

I thank E. Borra, D. Crabtree, G. Fahlman, Paul Hickson, D. C. Morton, M. Mountain, and D. Salmon for helpful comments on earlier versions of this work. A most informed referee's report clarified a number of subtle historical points.