Professor Roger Griffin and the Cambridge 36-inch telescope

R.F. Griffin’s career-long involvement with the 36-inch reflecting telescope, and its mirror & its dome, at the Institute of Astronomy, Cambridge, UK

First published here on 2021 November 24, and last modified on 2024 April 18
A 33-minute read

[PHOTO: 36-inch dome with Roger Griffin and telescope visible at open shutters: 43kB]

Above: Roger Griffin and the 36-inch telescope, seen through the open shutters of the dome. This photo was taken in the early 2000s from the roof of the 1951-vintage Workshop building, and shows the orange frame which holds the Coravel. (Photo: Roger Griffin collection)

An earlier version of this document was circulated privately on 2021 July 9 to assist members of the Institute of Astronomy (IoA) in familiarising themselves with the 36-inch telescope which few if any of them had had any occasion to visit, far less operate.

Roger’s death in 2021 February could have resulted in much of the content of this essay being permanently lost. However, I had taken an interest in my father’s stories, and had made notes of interesting/useful information relating to the 36-inch — only too well aware that otherwise precisely this eventuality would ultimately arise.

This document does not attempt to address the science which Roger (and others) conducted at this telescope, nor even his instruments themselves — both these categories (particularly the former) are well documented in peer-reviewed journals — but rather to record meta-science relating to the installation itself.

The following is factual (or stated as fact by Roger) and has been compiled from my notes, from emails, photographs and various items of paperwork which have come to light.

[PHOTO: Interior of 36-inch dome: 30kB]

Above: A photo of the observing floor, looking north-east from the top of the wooden staircase. Roger took this photo in early summer 1991 using his wide-angle (17mm) lens which he nicknamed “the estate-agent’s lens” [!]; he sought to illustrate the dome’s layout prior to the removal of his original spectrometer and its replacement with the Coravel.


[PHOTO: Observatory site and domes: 25kB]

Above: Roger took this north-east-facing view (among others) in late summer 1958 by climbing up one of two rickety wooden masts, used for an experimental radio-aerial in the field to the south of the Solar Physics Observatory (SPO). A guyline stabilising the mast cuts across the foreground. The 36-inch dome is nearest his camera. During the subsequent years, rows of conifers were planted between this vantage-point and the dome at Roger’s behest; the Workshop (mostly hidden behind the dome) and much of the SPO to its right were demolished; and then the Kavli and Battcock centres were built immediately beyond the 36-inch dome.

The Cambridge 36-inch reflecting telescope was built in 1951–55 by the firm of Sir Howard Grubb, Parsons & Co. at Newcastle-upon-Tyne. It is thought to be the largest telescope in the country by aperture. Its dome, 25½ feet (7.86 metres) in diameter(1), was built for a previous instrument(1) in 1912.

[PHOTO: Grubb Parsons logo plate: 17kB]

Roger Griffin saw the 36-inch telescope for the first time on 1955 October 27, when it must have been brand new. Barely 2 years later, on 1957 November 1, he used it for the first time, and almost immediately he set about making improvements to its spectrograph; within a decade he had developed his successful Photoelectric Radial-Velocity Spectrometer at this telescope, pioneering the cross-correlation method which subsequently became ubiquitous. By about 1980 he was the telescope’s sole user, and came to regard it as his own in all but name. By this time if not before, and thenceforth up to the end of his life, he took a proprietorial interest in the 36-inch.

[PHOTO: Roger Griffin at the 36-inch telescope: 41kB]

Above: Roger at the 36-inch telescope in 1979. Photo by Alan Griffin.

Roger first used the 36-inch telescope on the night of 1957 November 1, when he was just embarking upon his PhD.

He last recorded useful data with the 36-inch telescope 60 years later, on the night of 2017 November 12.

[PHOTO: The 36-inch dome by night: 18kB]

Above: A starry night at the 36-inch dome, admittedly with light pollution all too evident in the orange hue of the sky. Photo taken nearing midnight on 2004 March 16, while Roger was evidently observing in the South-West. Visible in the night sky (in consultation with an online sky-map) are Saturn (the brightest object at upper right), various stars in the constellation of Gemini, and Procyon / αCMi (the bright object nearest to the top of the dome) which was the subject of Roger & his wife’s first major joint project: the Procyon Atlas. The second bicycle is mine — I had ridden it from Oxford the previous day to visit my father.

Primary Mirror

[PHOTO: View down telescope tube at primary mirror: 25kB]

Above: A view straight down the telescope tube from the base of the tubular-framed section on 2012 July 26, when the mirrors had just been aluminised for what (as at 2022) would be the final time. Reflected in the primary mirror is (the photographer’s arm and camera, and) the secondary mirror and its image, which itself includes an image of the flat and of his camera. We are also seeing, besides sky, the inside of one of the dome shutters.

In the centre, suspended from either side, is the oval tertiary mirror: the ‘coudé flat’. Yes, that is a Post-it note affixed to its centre, where no starlight can fall because of shading from the secondary! Additionally, at the ten o’clock position on the rim of the telescope tube a dim orange-filtered lamp may be seen. The purpose of both the Post-it (scattering light) and the lamp (further illumination beyond ambient) is to provide some background illumination of the field of view, to facilitate manual guiding by making the slit (the point at which starlight is admitted to the spectrograph) visible to the observer.

The original mirror

The telescope’s original primary mirror was known to be an unsatisfactory piece of glass at the time that it was made, but in those post-war austerity days nothing better was obtainable; Roger described it as being made from a blue crystalline material that is only slightly analogous to glass, and noted that at the time of its provision the Director of the Observatories, Roderick Redman, estimated that it might serve for 15 years. The surface of the glass suffered from de-vitrification, decaying to become a mass of tiny cracks which scattered a lot of light. The original mirror still resides in its box, on the ground-floor area of the dome in the south-east.

The pyrex replacement mirror

By chance, Roger saw a listing dated 1988 November 30 of optical parts being offered for sale by the Fund for Astrophysical Research (F.A.R.), Inc, an organisation whose founding Scientific Director was Theodore (Ted) Dunham Jr. One such part listed was a pyrex 37-inch mirror blank, 7 inches thick, priced at AU$900.

Roger liaised with one of the F.A.R.’s trustees, Dr. Gerald E. (Gerry) Kron at Pinecrest Observatory in Honolulu, Hawaii, and duly bought the blank for the list price (which amounted to £591.88 Sterling) plus shipping charges of AU$835 (£418.84), both of which he paid out of his own pocket on 1989 June 30; the shipping actually came to only AU$755 and the surplus was returned to Roger. The mirror was packed in its own box of size 48 inches square by 17 inches deep, and came with its own lifting harness in the form of a steel band that encircles the mirror and is tightened with bolts; the whole package weighed 475kg. The pyrex disk was said to be many years old, probably about 50, and has a net weight of 200kg. The packaged blank was dispatched by Dr. Dennis Phythian of the University of Tasmania on 1989 August 25, and sailed from Melbourne on the ship Helen on voyage V230 on September 18, arriving at Felixstowe on/about November 8, subsequently arriving in Cambridge on 1989 November 20 at the hands of Kertainer Limited having attracted Duty and VAT of £184.33.

Following a hiatus caused by the telescope having no working instrument (q.v.) from 1991, Roger sold the new mirror-blank to Cambridge University for £1,843.00 in 1995 June. By this time it had been ground and figured very precisely near Cambridge by Jim Hysom of A.E. Optics Ltd.

Although the new mirror had not yet been aluminised, even so on Monday 20 February 1995 (as per Roger’s diary) it was possible to use it, in a manner of speaking, to look through the telescope — this relied on the small percentage of natural surface reflectivity of the figured pyrex disk. I helped Roger make this first use of his new mirror while I was visiting during my school’s half-term holiday: he looked at Capella from the ‘new’ (and as-yet un-instrumented) focus of the telescope at the north of the dome. This is what he is alluding to with the You know… in the second sentence of his 1996 email below.

To have it aluminised, Roger drove the mirror via Harwich both to and from Hamburg Observatory during  June of 1995; the work was carried out free of charge as a kindness, but would normally have cost of the order of DM1000.

[PHOTO: mirror-box and assorted junk: 28kB]

Above: The box containing the old mirror, with a defunct office chair and an equally otiose Electrolux fridge dumped atop its lid. The semicircular bands of the lifting apparatus for the new mirror can be glimpsed at far right; at far left may just be seen the mirror-trolley. These items are all in the south-eastern part of the dome at ground-floor level. (2021 May 14)

‘First Light’ of the new mirror

On 1996-11-17, Roger described in an email how his Coravel instrument was in a state of near-readiness [or so we all thought at the time], affording him an opportunity to try out his new mirror:

Date: Sun, 17 Nov 1996 22:47:04 +0000 (GMT)
From: Roger Griffin
To: Richard Griffin


    Oh yes, the new mirror has been in the telescope since quite soon after I got it aluminized.  You know I DID have it in and looked through the telescope B4 I took it to Hamburg, because I didn't want to do a Hubble with it - and when it seemd to give an image, quite near the expected place, I considered that it was probably all right.  In fact it does give a perfectly good image - in fact, when I opened up and set on Epsilon Cygni - the brightest star I have normally observed until now - while it was still broad daylight and the mirror no doubt cooler than the air, the image was extremely sharp and practically fell down the slit. Now I have got to get the flat aluminized - I think I can get it done at the RGO.  We used to have an 'exchange' flat - a 6-inch circular one in a wooden ellipse to pack it in place - but I don't think it will work now that I'm using the f/30 focus because the coude beam is bigger where it is reflected on the flat and needs the whole size of the ellipse (especially at low declinations).  The new primary mirror certainly looks beautifully shiny, in a way that the old one never did even when it was freshly coated.

(The doing a Hubble was a snide reference to the Hubble Space Telescope, whose primary mirror was discovered, only once the facility had been deployed into its Earth orbit, to have been incorrectly figured which resulted in severe spherical aberration. My parents were also dismissive of the Hubble Space Telescope because of the eye-watering amounts of funding which its delays hoovered up in the mid-1980s (claimed to be US$6 million per month), at the same time as the 100-inch Hooker telescope at Mount Wilson Observatory (which they used to great effect and considered to be the best telescope in the world) was closed through lack of funds.)


Setting-knobs and dial-gauges used in the adjustment of the telescope’s collimation are located beneath the primary mirror cell, and should generally be left well alone!

[PHOTO: Dial-gauge and Post-it notes on underside of telescope: 28kB]


The primary and coudé-flat mirrors, which when uncovered accumulate dirt that then adheres because of atmospheric moisture, were liable to need re-coating about every 2 to 3 years depending upon the quality of the coating. The secondary mirror, which faces downwards and so does not accumulate corrosive dirt, retained its 1954 coating applied by Grubb Parsons until 2008!

The replacement primary mirror was re-aluminised in 2002 and again in 2004 by Kendall Hyde at its site in Stroudley Road, Basingstoke. In late August of 2004, Kendall Hyde mis-handled the primary mirror and broke a chunk over 3 inches across out of the figured surface’s edge (another such, but smaller, damage diametrically opposite was apparently there already); they look terrible, but only lose between 1% and 2% of the mirror’s light-collecting power; they reduced their bill by £560 to the round figure of £1,000 by way of compensation.

[PHOTO: Damaged area of primary mirror: 29kB]

[PHOTO: Sliver of primary mirror on napkin, with ruler for scale: 32kB]

During the 2000s Roger had increasing difficulty finding anywhere that could aluminise such a large mirror, and considered (but rejected) driving it to Hamburg again. He had just the coudé flat aluminised in 2006 by Galvoptics, Basildon.

All three mirrors were re-aluminised in 2008 by Thin Metal Films Ltd which had taken over Kendall Hyde at its Basingstoke address, and again in 2012. As of 2022, the coatings are now a decade old and must be in an appalling state!

Removing and handling the mirror

In the south-west of the ground-floor area is a wooden holder which was built to accommodate the primary mirror (without its box) securely in the boot of Roger’s 1989 Mercedes estate-car, and which is designed to facilitate loading/removal using a standard fork-lift. Nearby should be rudimentary handling gear that came with the mirror: two semicircular bands that are bolted together round it providing a couple of lugs at opposite ends of a diameter, by which it can be lifted.

To lower the primary mirror out of the telescope it is supported by a system (which Roger described in a 2017 email but which I have not knowingly seen) that attaches to the undersides of the tines of the fork-mounting. The metal scissor-lift trolley which is used thereafter in raising, lowering and moving the primary mirror short distances is located in the ground-floor area to the east. It is no longer possible to remove the primary mirror from the dome using this trolley (see Demolition of “Warm Room” (2012) below). In connection with this, and with the perceived lack of sufficient knowledge in handling the mirror and removing it from the mirror-cell, in 2017 July on both 17th and 31st, Dr. Trevor Walker from Thin Metal Films in Basingstoke visited the dome with the departmental custodian to investigate how the mirrors could be extricated from the telescope and dome for aluminising.

[PHOTO: General view of fork-mounting: 29kB]

Above: A general view of the base-casting and the fork-mounting, with the bottom of the telescope tube visible at upper right; the black rectangle is the open flap giving access for the observer to remove the primary mirror cover, with the stool to facilitate this seen at lower right. The sidereal worm drive is encased in its oil-bath with a perspex lid just below the Setting Circle. (2012 July 26)

The box for the current primary mirror is missing, last seen in 2008 buried at the back of a shipping container at Lord’s Bridge.

Instruments and other Optics

[It is intended that a separate article will feature some of these in greater depth]

Original Radial-Velocity Spectrometer

[PHOTO: Roger Griffin by the observing desk: 33kB]

Above: Roger by his observing-desk in early summer 1991, when his original Radial-Velocity machine was about to be disbanded. Its output was the automated drawing by a biro in a pen-holder on a Brown-Recorder chart (nearest to right-hand edge of desk) which advanced at the rate of one inch per minute. (Photo: Roger Griffin collection)

In February 1966 Roger completed his original prototype Photoelectric Radial-Velocity Spectrometer, based upon an unfulfilled idea by Professor Peter Fellgett in 1953. It was used in conjunction with an f/18 secondary mirror which is no longer used (presumably stored in the dome somewhere?). The part of the instrument containing the cross-correlation mask and photomultipliers was located at ground-floor level, between the pillars which face a person walking through the double doors into the main part of the dome. Upon decommissioning in 1991 the Spectrometer was donated to the Science Museum who stored it at Wroughton in Wiltshire, item number 1992-1067.


In January 1988 Roger had placed an order with UCL for a greatly modernised Photoelectric Radial-Velocity Spectrometer, modelled closely on the 1979 CORAVEL instruments of Mayor / Baranne / Poncet which were installed at OHP on the Swiss 1-metre telescope, and at ESO. There were several significant difficulties with hardware, software, and interpersonal communication during the building and commissioning of Roger’s Coravel. Not until November 1996 was it first working, but even then further setbacks ensued until about 2000 by which time Dr Simon Mentha and Jon Schneider were engaged to bring fully into service the Cambridge Coravel, and to maintain and improve it. It uses the f/30 secondary mirror which, in conjunction with the tertiary coudé flat mirror, brings the starlight to a focus high up to the north of the telescope at the top of the orange frame; it is then brought vertically downwards by means of internal reflections within two right-angle quartz prisms cemented together. The initial image is re-imaged by a small lens, at the level of the white flange, at the position required for the Coravel.


The large brass eyepiece at the observer’s desk, with its 4-inch field-lens and plenty of eye-relief (welcomed by wearers of spectacles), dates from the 19th Century and was salvaged by Roger from the Sheepshanks 12-inch coudé refractor at Cambridge when it was being scrapped in the summer of 1959.

Opening-up Procedure

I am familiar with some but not all aspects of this procedure (the position of electrical main switches is not recalled). The observer needs first to open the dome shutters; then to clamber onto the west arm of the fork-mounting (using the rubberised matting glued there) to remove the secondary mirror’s cover (which contains a capsule of desiccant), to remove the large cover at the base of the tubular-frame section of the telescope tube, and to uncover the coudé flat. Then and only then should the observer walk downstairs and, with the aid of the low stool provided, open the hinged flap on the north face of the base of the telescope tube and lift/slide the cover of the primary mirror out of the mirror cell. Upstairs, a cover also has to be removed from the brass tube leading into the black cylindrical prism-module at high level on the orange frame on the north side of the dome. Closing-down would be a reversal of this sequence. It used to be necessary to grasp the metal handles on the large-diameter Setting Circle at the base of the polar axis, to align the correct Local Sidereal Time* with the graticule above the right-hand end of the worm-gear; electronic encoders, if they are operative, may have obviated this step.

* Local Sidereal Time could be read off a fine Dent regulator clock with deadbeat movement, mounted on the wall near the main doors, which Roger used to keep wound.

[PHOTO: Dome interior by day: 50kB]

Above: This view from 2012 July 26 depicts Roger in relaxed conversation with 4 visitors during a demonstration of the telescope. It also illustrates several of the features mentioned above, including the rubberised matting on the west arm of the fork.

Light Pollution

Roger had a row of evergreen trees planted to the south and south-east of the 36-inch dome in around 1970, to block out the streetlights of Madingley Road. He expended much time and effort endeavouring to fight against unnecessary and/or inappropriate lighting in the nearby area, including at an Athletics facility in Wilberforce Road, at the West Cambridge site, at the Madingley Road Park & Ride, and within the Observatories site itself.

[PHOTO: Screenshot of aerial imagery: 59kB]

Above: An oblique aerial view of the southerly part of the Observatories’ site, with north at the top. The 36-inch dome is in the lower centre of the picture. This photo must have been taken prior to 2008, as the Workshop (just north of the 36-inch) was then demolished followed a year later by most of the SPO to its immediate east. Roger’s row of trees is clearly visible, and a further row of them appears recently to have been planted. Imagery from Microsoft Virtual Earth.

Power Supply and Right-Ascension Tracking

Electricity supply in Cambridge was at 200V until around 1960, and the telescope’s equipment ran off 3-phase 200V from new (and still does). When Cambridge’s supply was changed to the present 240V, the electricity company disliked the idea of trying to convert the telescope’s slewing, setting/guiding and tracking motors to run on it; instead, they installed an auto-transformer to convert their new 240V supply down to 200V. It lives to this day on the left-hand side at floor-level just inside the door.

[PHOTO: Looking south within the original entrance-porch: 36kB]

Above: This view looking south from within the original entrance-porch clearly shows the auto-transformer. To the left is the double doorway from outside, and on the right is the double doorway leading through into the main body of the dome; through it can be seen part of the telescope’s fork-mounting, and the primary mirror’s removed cover leaning against one of the two brick pillars built to support the spectrographic equipment upstairs on the observing floor. (2012 July 26)

As new, the telescope’s Right ascension (RA) tracking equipment was supposed to generate its own 50Hz and run its synchronous motor off that, so as not to be subject to the frailties of the nominal 50Hz in the mains. But in practice, the tracking equipment tended to ‘latch onto’ the mains and not its own 50Hz. George Sisson, the MD of Grubb Parsons (and he of Sisson Fiddle fame, q.v.), arrived one day in person with a new pair of gears in his pocket and a few electronic components, and modified the drive so that it expected (and received) about 48Hz, which it was able to do without being bothered by the proximity of its frequency to the mains.

The original electronics occupied a large cabinet located under the stairs, nearly 2ft square and 5ft high. It contained power amplifiers (valves) the size of milk bottles, one for each of the three phases. The cabinet had 4 buttons and 2 lamps. The buttons had to be operated in sequence:

When Denis Beggs (Senior Technical Officer of the Cambridge Observatories) made a comparatively small transistorised power supply, he put it inside the same cabinet. Since it, too, could only run the synchronous motor but could not start it, the observer had to start on the mains as before. At least it worked at 50Hz, the original gears being restored in the gearbox leading to the worm-drive. But for reasons never understood (as Roger wrote, Beggs was singularly inscrutable and uncommunicative!) Beggs did not bother with phasing, and the user just pushed button 3 at any time — and if the phasing was more or less 180° off, then the fuse blew! Instead of arranging to check the phasing before transferring the load, Beggs simply bought fuses in boxes of 100.

Beggs died in 1993. A Health & Safety inspection in 2001 November revealed concerns about equipment that blew fuses so routinely, that possibly mixed the 240-Volt mains and the telescope’s own 200-Volt system, and whose metal casings appeared not to be properly earthed. Beggs’ device and its predecessor were removed, leaving the telescope’s tracking at the mercy of frequency-fluctuations in the domestic mains supply (requiring constant guiding-corrections every second or two). Roger eventually succeeded in persuading the Institute of Astronomy to commission a new solid-state device, which is still in use and is located on the North-East wall of the dome. It contains a digital oscillator whose frequency can be set to four decimal places; it takes in the 3-phase mains supply, produces 3 more in anti-phase with those (so that there are six supplies each 60° from the next); and thus it produces quasi-perfect 50Hz output simply by switching between the 6 phases as required.

Right-Ascension Tracking Errors

The sidereal tracking-drive had a series of periodic errors arising, I believe, from manufacturing defects in the worm-drive. The sidereal worm-gear takes 80 seconds to revolve. In around 1966 Roger and the then Director, Redman, investigated the periodic errors — Roger watched a star-image in the eyepiece and called out “It’s going off to the left”... “now it’s off to the right...”, while Redman drew a vector diagram of the errors’ magnitude and direction against the period of the worm-gear’s rotation.

George Sisson (q.v.) produced a device which indirectly coupled the RA worm-gear to the worm-gear which drove it; the device resembled a pair of clutch-plates with an adjustable disk on studs and rollers. It was known as the Sisson Fiddle: it resolved the periodic error, and it is still in situ.

I have encountered a contemporary handwritten note (below), in Redman’s own hand, regarding “Adjustment of Sisson Fiddle” (not that this ought now to be required); it says,

If the total range of the guiding error is x seconds of arc, and the maximum excursion of the star-image to the west comes at phase θ, measured from a zero when the main worm shaft locking pin is vertical and the 1 is on top, the required adjustment is:

x sin θ turns down
x cos θ turns north.

[PHOTO: Handwriting on notebook page: 41kB]

Telescope Motions

The telescope is fitted with no limit-switches of any kind. It is entirely the observer’s responsibility to avoid collisions (whether in slewing, ‘R.A. Free’ (q.v.), setting/guiding, or tracking) between telescope tube and the edge of the hole in the floor, the staircase, the observer’s desk, the orange frame, or any humans or other obstructions placed in the dome. Likewise there is nothing (except prior collisions involving the telescope tube) to prevent the mirror from being tipped out of its cell.

A zone between approximately 76° and 81° North is unviewable with the current coudé arrangement, as the top ring of the telescope tube blocks the light-path from the coudé flat.

The slewing controls and the slewing/guiding switches were originally mounted within a pedestal which also contained the dials displaying the telescope’s position. When the observer’s desk was repositioned for the installation of the Coravel, the pedestal’s components were redistributed; the slewing control-buttons and switches were then mounted on the the edge of the desk.

Roger repeatedly reported difficulties with the reliability of the slewing and setting/guiding motions, both RA and dec — certain electrical relays would either fail to operate reliably, or become ‘stuck in’ which would snatch the star away from the field of view. In the guiding-paddle on the observer’s desk, the microswitches under the buttons were a regular source of trouble. It appears that Roger and the workshop staff were engaged in an ongoing battle against moisture impairing this equipment.

The RA motion has an unusual facility, a key-switch marked ‘R.A. FREE’ (located below the RA buttons and switch, q.v.) which when operated disengages the slew-mechanism (and its brake), allowing the observer to move the telescope and its mounting by hand about the polar axis. Care must of course be taken to avoid allowing the telescope to ‘run away’; likewise the key-switch must be normalised only once all movement has completely ceased.

[PHOTO: View of (and down) the top of the telescope tube: 54kB]

Above: The top ring of the telescope tube has 4 finials around it as may be seen, though their purpose is not known to me. The metal quadruped-‘spider’ holds the secondary mirror in its cell, which incorporates the focussing-alteration motor that is visible. Also visible is the orange frame which carries the Coravel, and part of the vertical light-tube which gathers the starlight to a focus high up in the dome to the north. (2012 July 26)

An unrelated quirk of the telescope is that, at certain positions, the telescope balances rather too well and can ‘flop around’ on the ‘play’ in the RA and dec motions. In such cases, Roger was known to hang a chair on one of the finials at the top of the telescope tube, to provide ‘preload’ and so suppress this unwanted motion.


As previously hinted, the dome could be quite a damp environment. Three powerful fans in the ground-floor walls expel air at a flow rate of 6,000 cubic feet per minute, drawing ambient air in through the dome shutters. In fact, though, these fans were part of a 4-pronged approach to improving the ‘seeing’, the other prongs being as his thesis(1) describes, (1) lining the centrepiece of the telescope with thermally insulating material; (2) installing three fans at the back of the mirror cell to draw air down the telescope tube and around the mirror, thereby preventing convection currents above the mirror; and (3) painting the whole of the outside of the building and dome white to reduce daytime heating..

Sometimes Roger found it necessary to remove dew from the primary mirror using a hair-dryer, which is no doubt to be found in the dome. A fan-heater was also present at times, as it was necessary periodically to change over a cartridge of desiccant that lived within the Coravel housing (it could then be baked in a small oven beneath the stairs), but even opening the casing to change the desiccant could risk dewing the internal parts of the instrument during the dampest conditions.


On the longest night of 2009, December 20-21, Roger observed from 1640 to 0720! The weather had been snowy/icy for the past week or so. Rotating the dome, which was done by turning a handle, was proving to be very difficult, presumably because of ice blocking the guttering where the moving part of the dome comes down into it. One particularly difficult sector he found easier to bypass by turning the dome all the way round, rather than trying and force the dome through it.

Then, on 2009 December 22 he found the rotation of the dome was totally frozen up, stuck at 30 degrees West of South. He even borrowed someone from a nearby building and they tried together but it wouldn’t move. He also rigged up a lever that would allow him to apply the force of ten men, but it still didn’t budge. So he observed through the dome in the position that it was in, until 0315hrs when he’d done everything useful from that point and gave it up as a bad job!

The “Warm Room” (1990)

As built, the circular dome structure had a single-storey brick ‘annexe’ adjoining its north aspect, some 5.1 metres long and 3.3 metres wide, which comprised the dome’s entrance-porch and a plate-store cupboard.

In winter 1989 / spring 1990, as part of the re-equipping of the telescope with the Coravel instrument, the existing single-storey ‘annexe’ on the northern side of the dome was extended, nearly doubling its length to 9.3 metres. This extension housed a computer-room, and such novel creature comforts as running water and a toilet.

[PHOTO: Foundations being dug: 28kB] [PHOTO: Concrete floor-pad laid: 30kB]

Above: During the winter months, work began on building the “warm room”. Photos by Roger Griffin.

[PHOTO: : 36kB]

Above & below: As winter gave way to spring 1990 and with the brickwork complete, the roof was fitted; a concrete apron was laid alongside the full length of the building. Photos by Roger Griffin.

[PHOTO: : 40kB]

The computer and associated electronics for the new Coravel then ‘on order’ would live in this “Warm Room”; once the star had been acquired and the auto-guider enabled, the observer could then — as at the Observatoire de Haute-Provence (and indeed at most observatories) — retreat from the ambient (often very cold and windy) observing-floor of the dome to the relative comfort of this new room.

[PHOTO: Interior of “warm room”: 33kB]

Above & below: The interior of the “Warm Room”, including a general purpose computer and storage, and creature comforts including a sink and a heater, as well as computer-terminals and various items of electronics relating to the Coravel. Unsurprisingly to those who knew him, the shelf at high level in the second photo bears all of the many hallmarks of being the fruit of one of Roger’s many carpentry projects. Photos by Roger Griffin.

[PHOTO: Interior of “warm room”: 40kB]

On at least three occasions (on 2009 October 13, during 2011 April, and on around 2017 November 19), the lead roofing on the extended porch / warm room (or in the latter case its truncated remnant) was repeatedly stolen.

[PHOTO: Dome with roof sheeted over: 33kB]

Above: A tarpaulin covers the roof of the dome’s ‘annexe’ whose roofing-lead had been stolen. The acreage of cleared area in the foreground is where the SPO and Workshop had stood until recently. The theft occurred one week after the demolition contractors had finished their nearby work. (2009 November 20)

[PHOTO: Dome with roof sheeted over again: 39kB]

Above: Same problem, same tarpaulins, different occasion. Roger said it had been stolen about a week before our visit, and that the security-camera which he’d been assured would now be trained upon it had in fact not been. (2011 May 1)

[PHOTO: 36-inch dome, western face: 38kB]

Above: The western face of the 36-inch dome, seen on 2012 July 26 with an uncanny absence of buildings behind it (the SPO and Workshop having long-since been removed). The join between the original ‘annexe’ and the “warm room” is apparent in the brickwork, and the position of the vent-pipe from the toilet demonstrates that this was built into the space previously occupied by the original plate-store.

Washing the dome

[PHOTO: Roger undertaking dome-washing: 33kB]

Above & below: Roger has just washed one panel of the dome’s skin by way of a test, where the result compares tolerably well with a sheet of white A4 paper. It is apparent that this task has become considerably overdue! He must have walked out through the open shutters to stand on the roof of the ‘annexe’ to access the turning part of the dome, which (thinking about it) must then have been turned by an assistant who must also have taken this photo with Roger’s camera. Photos: Roger Griffin collection, circa 2010.

[PHOTO: Roger undertaking dome-washing: 29kB]

Kavli and Battcock

In 2008 the Departmental Workshop was demolished, with the Von Klüber House and much of the SPO following it in 2009, as part of the construction of the Kavli Institute of Cosmology, sited well to the north of the 36-inch in an area once occupied by the gardening staff’s shed and (many years ago) the Newall telescope.

[PHOTO: Aerial imagery of Observatories site: 41kB]

Above: This aerial view from 2008, from Bing Maps, shows the Kavli building being constructed well to the north of the 36-inch dome, with the contractors’ temporary access-road leading from a new entrance on Madingley Road, across the SPO lawn and through the site of the Workshop. The SPO was itself demolished soon after.

By 2012 a previously-threatened plan to build the Battcock Centre very close to the 36-inch dome was coming to fruition. For reasons never satisfactorily explained, the Battcock building — which was to form a quadrangle with the existing Kavli and Hoyle buildings — was to have its main axis oriented at an angle to the others in a manner which brought its footprint much closer still to the 36-inch dome: so much so that part of the dome would have to be demolished.

[DRAWING: Roger’s annotation to plan of existing buildings: 66kB]

Above: Roger annotated a plan of the then existing buildings, showing how the obvious place for another building would both retain the rectilinearity of the quadrangle, and avoid requiring the partial demolition of the 36-inch dome. Note for fellow historians — two of the printed captions on the plan are mistaken: “Thorrowgood” should read “Northumberland”, and “Workshop” should read “SPO” or maybe “Outreach Centre”.

[PHOTO: Aerial view of Observatories site showing Battcock Centre: 103B]

Above: The position of the Battcock Centre as actually built, as seen on aerial imagery from Google Earth dated 2021-06-13.

Demolition of the “Warm Room” (2012)

By late February of 2012 it was becoming apparent to Roger that the Battcock building’s design would require the demolition of the 36-inch dome’s 1990-vintage “Warm Room”, and additionally half of its original 1912-vintage entrance-porch, leaving only 2.5 metres’ length (or around one-quarter of the building) extant.

[DRAWING: Scan of Roger’s alternative plan: 51kB]

Some 5 years previously, Roger had seen a plan that showed such a development overlapping the dome, and of course he had complained; the then Departmental Secretary, Paul Aslin, had assured Roger that he should not worry as there was no money for that hypothetical building anyway, and said that if they “had to” demolish the room that he’d had built onto the dome they’d “just rebuild it in another orientation”. Roger produced plans for such an eventuality (above), but this undertaking was not honoured; and his subsequent attempts to discuss the matter with those in charge were stonewalled. In desperation, Roger sought help from the Vice-Chancellor of Cambridge University (his letter, which includes reference to an alleged requirement for a fire engine to have access between the dome and the new building, also contained the annotated plan shown above), but this had no effect either.

[PHOTO: Lynden-Bell and Griffin at Heathrow: 34kB]

Instead, on 2012 March 28, while Roger was known to be incommuni­cado in the USA for filming of the docu­mentary Star Men with former IoA Director Professor Donald Lynden-Bell (see illustration of their return through Heathrow Terminal 5 at lunchtime on 2012 April 1, above), an admini­strator emailed Roger giving three proposals for incorporating a loo into the rebuilt entrance-porch where there was not room for it.

[DRAWING: Options for demolishing/rebuilding dome ‘annexe’: 38kB]

Above: The three options offered for demolishing and rebuilding the dome’s ‘annexe’. All of these were drawn by somebody who had no comprehension of the activities which the building needed to be capable of supporting.

Hearing no reply from Roger within two days, Option 3 was then unilaterally selected without an understanding the ramifications thereof. The resulting design reduced the width of the original entrance-doorway, which used to have double doors, to just a single-door’s width.

The new doorway is too narrow for the 36-inch primary mirror on its trolley to be removed from the dome, thus greatly increasing the risk and complexity of any attempts to re-aluminise it — none has been attempted since.

The demolition took place during 2012 September.

Roger attempted to continue observing during the start of these works — for his desire to gather data from ‘his’ stars far outweighed other considerations. In contemporary email (reproduced below) he describes his blackleg observing, and also makes unremarked references to a particularly euphemistic meaning of the verb to modify which had been rather callously coined by a proponent of the demolition scheme:

Date: Tue, 18 Sep 2012 23:47:07 +0100 (BST)
From: Roger Griffin
To: Richard Griffin

    Thank you so much for two messages; my apologies for not responding till now to the first one.  I have been very busy, not only with getting everything out of the room in the dome […] and also observing.  There have been only five nights in September when I have not been observing, though on one other I was clouded out when I had got only the first standard star so I got nothing useful.  Indeed, yesterday was the day when they came to modify the dome.  They had done nothing except bring a portakabin by last night, so I took the opportunity to observe again, although it has been getting more difficult since they broke the network connection a week ago, so I couldn't watch what is happening from my office or even (as far as the autoguider is concerned) from downstairs in the dome.  And there has been no phone to the dome, because the phone works through the network now, and the meteorological information is no longer displayed.  I was asleep today until the afternoon, of course.  A great builder's fence, sorry, I mean a great fence set up by the builders, sorry, I mean contractors, has set itself up.  I found weak point in it and went to see what had happened in the dome: they have torn out the electrics in the room and turned off taps with a monkey wrench so they can not be turned on again, but apart from having seemingly turned off the autoguiding computer the rest of the equipment was working, so I am doing a spot of blackleg observing now.  I *hope* that the telescope etc are behaving themselves, because I've no way of checking on them.  I have two new stars that need long integration about this time of night, and one 'old' one at a mega-exciting phase.  I shall be much upset if I am prevented from observing for the next six weeks, which are critical to it.


    I must go back to the dome, quite likely something bad will have happened by now.

[PHOTO: Dome just prior to demolition works: 46kB]

Above: The dome just prior to works commencing. Photo by Roger Griffin.

[PHOTO: Dome during demolition work: 49kB]

Above: The shell of the “Warm Room” is partially standing, and the verdigris on the dome roof seems to have got worse as if in sympathy. Photo by Roger Griffin.

Date: Mon, 1 Oct 2012 18:46:21 +0100 (BST)
From: Roger Griffin
To: Richard Griffin

    The modification itself has been completed, and at last noticing they were building a wall where the breeze-block partition between the airlock and the loo used to be.  I took some pictures of the dome from various angles before they started, but I have not kept an hourly cine-film of what they have been doing to it.  I am forbidden to observe until they have finished, which incredibly is billed for nearly the end of this month.  It is particularly annoying to me to lose the opportunity to observe the star whose velocity plot is attached, which has a very sudden rise in velocity.  I have not been allowed to observe since it was at the minimum - you can see it was very well observed on the approach to there - and it is now already up to about -15 and by the time they let me try to observe again it will be where the arrow points - no good at all.  It won't come round again when the star is accessible until 2019, and at my age such a delay doesn't bear thinking about!  And it is so annoying that there is no *reason* why I should not observe except a prohibition.

[DRAWING: PNG version of PDF plot of HD 218925: 28kB]

[PHOTO: Truncated dome in building-site: 64kB]

Above: The newly truncated dome, complete with narrowed doorway, with a foreground of groundwork taking place for the Battcock Centre. (2012 October 31)

[PHOTO: Building site viewed through dome shutters: 68kB]

Above: With the dome rotated to just west of north, the view through the shutters is of the roof of the truncated ‘annexe’, form-work for the casting of reinforced concrete structure, and a Portakabin. Photo: Roger Griffin.

[PHOTO: Dome with new building immediately adjacent: 60kB]

Above: The Battcock building has taken shape behind the truncated 36-inch dome which gives every indication of being actually within the building-site. Photo: Roger Griffin.

[PHOTO: A bulldozer on a building-site beside the 36-inch dome: 57kB]

Above: I am certain that Roger must have found this photo-opportunity upsetting and irresistible in equal measure — a bulldozer seemingly having just demolished the room built for his Coravel; his ever-present bicycle is propped against the Heras fencing to the left.

[PHOTO: Newly narrowed doorway to dome: 29kB]

Above: Roger standing in the dome’s newly narrowed doorway, through which it would be somewhere between ‘tiresomely awkward’ and ‘impossible’ to remove the primary mirror. His expression does not hide his dissatisfaction with this state of affairs! The extractor-fan for the ill-advisedly retained toilet may be seen at the top-right corner. 2013 May 10.

Dome Re-cladding

For bona fide scientific telescopes, it is essential that the dome and its contents be maintained at temperatures as close as possible to those of the night-time. Otherwise, thermal distortion of the mirrors and other equipment, and turbulent convection-currents (“seeing”), impair the ability of the telescope to focus starlight.

Roger, with decades of experience in the art, somehow just knew what to do to optimise the telescope’s “seeing”, by stabilising the air in the dome, particularly any convection-currents immediately in the light-path of the telescope. He might use one, two or all three of the Dome Fans (q.v.), according to his reckoning of the prevailing conditions. To avoid upsetting favourable air-currents within the dome, he was quite capable of selecting stars from his observing-programme which were attainable without turning the dome — sometimes for several hours at a stretch!

The hemispherical rotating section of the dome had been clad in copper sheeting since around 1950, replacing previous papier-mâché; the copper was soon painted with white paint whose pigment was Titanium Dioxide. Such paint is black in the infrared while being white in the visible; so while reflecting daylight and sunshine, it radiates hugely at ambient temperature — and so it does not heat up in sunshine but instead cools itself by radiating away what heat it has got, thereby usually keeping itself (and whatever it is painted onto) actually below the ambient temperature. There are other white pigments but they do not have the same property.

[PHOTO: Dome covered in scaffolding: 39kB]

Above: The dome during re-cladding work. The sky may be seen through the bare ribs of the upper part of the dome structure. 2017 April 19.

On 2017 January 30, the 36-inch dome was taken out of use for a nominal 12-week programme of refurbishment, which was to include the removal of the copper sheeting and its replacement with a plastic cladding. The purpose of this re-cladding was never satisfactorily established; it may have arisen from a mis-understanding as to the source of rainwater-leaks, which were actually around the shutter and so required rectification to weather-seals, not to the cladding itself. Nor, in spite of Roger’s repeated written requests, was the dome cladding of a material which had equivalent thermal properties to Titanium Dioxide pigment; instead, the plastic cladding was fully covered by a self-adhesive film of light-grey Bauder THERMOFOL [sic] which has no specific sunlight-reflecting properties whatsoever, nor was it (nor perhaps could it have been?) painted. The completion of these works was much delayed, the telescope not becoming available until around 2017 July 13. After this date, I witnessed a daytime temperature within the dome in excess of 30°C.

[PHOTO: Dome and Battcock Centre: 61kB]

Above: The re-clad truncated 36-inch dome with the Battcock Centre building immediately adjacent. 2021 May 27.

Roger was either not involved in the decision-making processes for these works, or his (by then) increasing cognitive impairments precluded his effective participation in them. In any case the combination of:

all resulted in the re-cladding hiatus marking the effective permanent cessation of his data-gathering at this instrument.

[PHOTO: Logbook entry: 34kB]

Above: Roger’s entries in his A5 spiral-bound observing logbook come to a sticky end on the night of 2017 January 28, after completing an observation of star H 70 (better known as ε Tau — Epsilon Tauri was one of Roger’s fundamental Standard Stars by which he calibrated his radial-velocity observations) in the Hyades star-cluster, and then aborting (no doubt because of cloud) an attempt to observe HD 52633 which he’d begun at 22:04hrs. The night was evidently not good meteorologically, and there’s a reference to phoning Dr. Simon Mentha for real-time assistance to resolve an issue he’d had while trying to observe Hyades star H 75 (HD 28363) a little earlier.

This was the last “fine” (!) night prior to the re-cladding hiatus, and Roger appears to have misplaced this logbook in the intervening months.

Following the re-cladding hiatus, with considerable assistance he took (potentially?) useful observations on just one night: Sunday 2017 November 12.

[PHOTO: Roger at his desk in the dome: 28kB]

Above: Roger in his favourite place: at the eyepiece and controls of the telescope which he used for most of his life. The screen carries the interface for the Coravel and is showing a typical trace complete with pronounced ‘dip’. Photo: Roger Griffin collection, photographer unknown, please get in touch.

Footnote (1) — According to Roger’s dissertation Direct photoelectric measurement of the intensities of features in stellar spectra, submitted in February 1960 in support of his candidature for a Fellowship at St. John’s College Cambridge, the dome is 27 feet in diameter (and taking a measurement off a scale cross-sectional diagram of the (interior surface of the?) dome yields a figure of 26½ feet); his text says,

The Cambridge 36-inch reflector was manufctured by Grubb Parsons and installed in 1955 in the 27-foot dome hitherto housing the old Common 36-inch, the latter instrument being returned to its owners, the Science Museum. The new telescope was designed for use either at the prime focus (f/4.5) or at a coudé focus (f/18 or f/30).