1. Why Big Telescopes Switched from Lenses to Mirrors
The first telescope used by Galileo in 1609 is said to have had an objective lens with an aperture of about 3.7 cm [3]. Refracting telescopes then grew larger, reaching about 102 cm with the Yerkes telescope in the late 19th century. This is described as the largest refracting telescope in use [3]. There are two likely reasons why growth stopped there.
One is color blurring. A single lens bends light differently depending on its color, causing false colors at the edges of stars. This is called chromatic aberration [1]. In 1668, Newton in England created a reflecting telescope, using a concave mirror instead of a lens to solve this flaw [1]. An explanation for a 1924 replica in a science museum also states that Newton chose a concave mirror to avoid false colors [2]. Since mirrors only reflect light, they do not bend colors differently.
The other reason is weight. A lens must let light pass through, so it can only be supported at the edges. As it gets larger, it sags under its own weight, distorting the image; this is considered the limit [3]. A mirror, however, can be supported across its entire back surface. So, did anyone in Edo-period Japan make a reflecting telescope?
2. Kunimune Ikkansai, an Edo Gunsmith, Builds a Metal Mirror Telescope
In 1834, Kunimune Ikkansai (1778–1840), a gunsmith from Kunitomo Village in Omi Province, made a reflecting telescope. The magnification was about 70 times. This telescope, recorded as a national cultural heritage, has both its primary and secondary mirrors made of a special alloy of about 63% copper and about 37% tin [5].
Polishing metal into a mirror, rather than using glass, was likely a skill unique to gunsmiths. It is said that Ikkansai made about six reflecting telescopes, and four still exist today. All are Gregorian telescopes, which use a parabolic concave mirror as the primary mirror [6].
Reports say that the primary and secondary mirrors remain in usable condition even today, about 200 years later [6]. It is said that Ikkansai used this telescope to observe sunspots for the first time in Japan with a self-made reflecting telescope, and more than 200 sketches remain [6]. Since the number of sketches and days of observation vary slightly depending on the source, we say “more than 200” here.
However, metal mirrors had a weakness: they tarnished easily. How was this overcome?
3. From Tarnishing Metal Mirrors to Silver-Coated Glass Mirrors
In 1857, Léon Foucault in France made a reflecting telescope using a silvered glass mirror [4]. He polished a glass sphere and applied a thin layer of silver on top. Unlike metal mirrors, which tarnished easily, if the silver tarnished, it could be reapplied to restore the original performance because the shape remained in the glass.
However, some sources say that Steinheil was the first to make a silvered mirror in 1856, so whether Foucault was truly the “first” depends on the source.
With the glass holding the shape and the thin silver providing the shine, this division of labor made mirrors easier to repair and reuse compared to metal mirrors. The next problem was that when made large, the glass shape would distort due to weight and heat.
4. Why a 5-Meter Mirror Is Too Heavy: The Hale Telescope and Okayama's 188 cm Mirror
The main mirror of the Hale Telescope at Palomar Mountain in California has a diameter of 200 inches (about 5.1 m) and was completed in 1948. Until 1993, it was the largest effective telescope in the world [7].
This mirror was commissioned from Corning in 1929 and made of Pyrex, which expands little with heat. The back was shaped like waves (waffle-like) to reduce weight [8]. Choosing glass that holds its shape despite temperature changes and lightening the back were the two tricks for making a large single mirror.
In Japan, the 188 cm reflecting telescope at the Okayama Branch Office of the National Astronomical Observatory of Japan was completed in 1960. With an aperture of 188 cm, it was ranked 7th in the world at the time and called the “finest telescope in the East.” It was made by the Grubb Parsons company in the UK [9]. The primary mirror is made of Pyrex, with an effective diameter of 1.88 m, a thickness of 0.27 m, and a weight of 1.7 tons [9]. Dividing the diameter by the thickness gives about 7, so the mirror’s thickness is about 1/7 of its diameter. The mirror was kept thick to maintain its shape.
So, how does a telescope with a thickness reduced to about 1/41 of its diameter maintain its shape?
5. How the Subaru Telescope's 20 cm Thick Mirror Is Adjusted by 261 Supports
The main mirror of the Subaru telescope has an effective aperture of 8.2 m, a thickness of 20 cm, and weighs 22.8 tons. It is made of ULE glass (ultra-low expansion glass). It uses a technique called “active optics,” where 261 actuators (devices that push and pull) support it from the back and continuously adjust its shape. The precision of the mirror surface is 0.014 microns [10].
Comparing the numbers, Subaru’s diameter is about 41 times its thickness (8.2 ÷ 0.2). The Okayama 188 cm mirror is about 7 times (1.88 ÷ 0.27). Looking at thickness as a ratio to diameter, Subaru is about 1/41, while the Okayama 188 cm mirror is about 1/7. The ratio of effective diameters is about 4.4 times (8.2 ÷ 1.88), and the light-collecting area is about 19 times (4.4 squared). Whether the ability to make it so thin is determined solely by active optics is not explicitly stated on the Subaru official page. Here, we stick to the certain fact that it is “supported by 261 actuators.”
The mirror surface is covered with a thin film of aluminum, which is re-coated every few years to maintain high reflectivity [11]. Foucault’s idea of “reapplying silver” continues today.
Still, a single mirror has limits. What should be done if it exceeds 10 meters?
6. Splitting a Mirror into Many Pieces: The Keck and Seimei Telescopes
The main mirror of the Keck Telescope in Hawaii consists of 36 hexagonal mirrors, each about 1.8 m wide, 7.5 cm thick, and weighing about 0.5 tons, arranged like a honeycomb. 168 sensors and position-adjusting actuators at the edges correct misalignments twice per second [12].
The proposal for this method was that a single 10-meter mirror would require a complex support structure and huge costs, but if segmented mirror technology is established, there would be no essential limit to the size of the reflecting surface [12].
In Japan, the Seimei Telescope at Kyoto University’s Okayama Astrophysical Observatory (main mirror aperture 3.8 m) is a segmented primary mirror made of 18 fan-shaped mirrors. In July 2024, it was confirmed that the height of each segment was aligned to a precision of about the wavelength of light. This is the first step toward aligning the segments in phase [13]. Before this, observations were done in a state called a “light bucket,” where only the focus position was aligned [13].
We see two paths to large mirrors: a thin single mirror, and mirrors split and arranged.
7. Measure a Mirror's Diameter and Thickness at Home
What you can do today: Measure the diameter and thickness of a hand mirror at home with a ruler. Divide the diameter by the thickness to find out “how many times the diameter is the thickness.” For the Okayama 188 cm mirror, the diameter is about 7 times the thickness, and for Subaru, it is about 41 times. Comparing these helps you realize how thin Subaru’s mirror is.
Also, if you look at your face in the concave side of a spoon, it will appear upside down. This is a familiar example of a concave mirror.
Official information: The National Astronomical Observatory and the Subaru Telescope official pages have photos and explanations of the mirrors. You can open them from the source list of this article.