Amateur astronomer Mark McCarthy recreates the Herschels' legendary "Night of Discovery"


This month marks the end of the first year of my stargazing adventure. Almost everything I look at is new, and a thrilling discovery to me. So when I saw an internet forum post referencing the S&T article on William Herschel’s “Night of Discovery” of April 11, 1785, during which he discovered 73 galaxies, I thought to celebrate by replicating his observation session, known as the “Herschel Sprint.” As an extra challenge, I decided to approximate Herschel’s experience as closely as I could.

Spiral galaxy NGC 4911 against star field
William Herschel discovered barred spiral galaxy NGC 4911 on April 11, 1785.
NASA / ESA / Hubble Heritage Team (STScI / AURA)

S&T posted a list of all the objects to be observed on their website. They suggested “purists” could make the sweep more authentic by using the drift method — fixing the telescope at meridian and only moving it 2-3° in altitude, letting the objects drift by as Herschel was forced to because of his mount. One could approximate his view by using an older style eyepiece with 45-50° apparent field of view (AFOV) at 150×.

But why stop there, I thought? His telescope used an 18.7-inch speculum mirror, which (if Wikipedia is to be believed) has 66% of the reflectivity of a modern mirror, equating to an approximate effective aperture of 12.5 inches. I happen to have a 12.5-inch f/7 scope! I could use the 15-mm RKE eyepiece from my Astroscan to match the 45° AFOV, at 148x in the 12.5-inch.

Best of all, I had previously observed only five of the objects on the list. These five objects (NGC 3245, 3486, 3504, 3277, and 3414) were observed on February 16, 2015 using a 20-inch telescope. Unless I looked in my notes, I had no memory of what they looked like; besides, they would appear significantly dimmer in my 12.5-inch scope. So, 68 of 73 objects would be completely new to me, and 5 of 73 would be “practically” new.

Incidentally, I’m now one month younger than Herschel was in 1785, so we have our ages in common too! I remember reading in a 19th-century biography that Herschel intensified his deep-sky observing when he found himself “on the wrong side of 45,” as Caroline Herschel is said to have put it. This more or less explains the drive behind my observing.

I took care to preserve my ignorance, such as it was, by not doing any new research about Herschel’s night or the objects to be observed. I downloaded the spreadsheet list of objects from the S&T website, but deleted all of the columns except the NGC number, RA, and Dec. — other columns contained some hints, like “low surface brightness,” so I avoided them.

I wanted some way of confirming the object I was looking at during the sweep. At first I thought of navigating by star-hopping in the eyepiece; so I photocopied the relevant pages from Uranometria 2000.0 and traced lines between each object on the list to plot my course.

But I soon realized I could easily get lost; estimating how many degrees to move in the eyepiece over a several-hours-long session would be confusing. Then I remembered from my prior reading that Caroline Herschel, as she took dictation, had next to her an accurate clock set to sidereal time, and the telescope had a contraption using ropes, pulleys, and a barrel top marked in altitude degrees. As she took dictation, she could note an object’s right ascension from the clock and the declination from the scale on the barrel. I didn’t think it would be cheating to use modern equivalents: a digital clock and a digital inclinometer. I could find a star at 0° declination that was also preceding the first object on the list. When this star reached meridian, and with it centered in the eyepiece, I could set the time on the clock to the star’s sidereal time and set the inclinometer mounted to my tube to zero. Thus calibrated, I could move my scope in altitude only to the first list object’s declination and wait for it to pass through. So long as I didn’t move in azimuth, these guides would remain accurate enough throughout the session and help me find the objects.

The S&T spreadsheet showed right ascension in tenths of minutes, and declination in seconds of arc; I converted these into minutes & seconds for right ascension, and tenths of degrees for declination, to match what I would see on the digital clock and the inclinometer, and printed it out to use at the telescope. I kept my Uranometria charts as back-up, but did not refer to them to find the list objects.

This would be a severe test of my visual descriptive skills, which — since the objects pass through the FOV in less than a minute — would require great economy of perception and expression. There wouldn’t be enough time to dwell on an object, search for fine detail, or write and sketch my observation as I normally do, but I did have a digital voice recorder that I velcroed near the focuser for dictation.

What would run through my mind as the pageant unfurled before me? All that was left was to wait for a cloudless, moonless night.


I arrived at Fremont Peak Observatory Association's site at Fremont Peak State Park, California, at sundown and was observing soon after. I used the 15-mm RKE from the beginning, because I wanted to become accustomed to its FOV and how objects would appear.

I ran through some H400 objects, and could tell right away the night ahead would be more challenging than I thought. My February session at the Peak under similar conditions was with my 20-inch, and I recall all my observations spoke of finding objects with direct vision, with averted vision helping to reveal some fine detail. This time, objects were still bright enough to be found but averted vision (AV) was more necessary to bring out more of the object, such as a galaxy’s halo or mottling.

Only a handful of the Sprint objects are plotted in my Pocket Sky Atlas — which is targeted for small and medium size scopes and includes all the Messier, Herschel 400, and Caldwell objects. I was still able to find the H400 objects (mostly in Leo’s hindquarters) perfectly well and enjoy them, but the comparative dimness with the 20-inch was a foreshadowing of how the Sprint session would go.

William and Caroline Herschel discovered some 20 new galaxies in the heart of the Coma Cluster (Abell 1656).POSS-II / CalTech / Palomar Observatory
William and Caroline Herschel discovered some 20 new galaxies in the heart of the Coma Cluster (Abell 1656).
POSS-II / CalTech / Palomar Observatory

I used Stellarium beforehand to find a preceding, 0° calibration star that was bright enough for me to find. I used HIP48413 (RA 9h 52m 59s, Dec 0° 00′ 13.6″), which according to Stellarium would reach meridian at 11:14:35 p.m. PDT on 3/18/15. It was an easy find off Iota (ι) Hydrae. As it neared meridian I kept it in the center of view and monitored my watch; at the right moment I set the inclinometer to zero, and set my clock (an alarm clock with red numerals) to 9:53. Then I moved the scope up to 27.7° to wait for the first object, NGC 3196.

But while I was waiting the inclinometer’s backlight went off. I thought the unit itself turned off; when I pressed the “on” button, the unit reset zero to default bubble level zero! I went back to my calibration star and reset to zero again, but since I lost the sidereal time I skipped NGC 3196 and went to NGC 3245, which I found by star-hopping with the chart. Fortunately the declination angle matched what it should be. I held the galaxy in view until the sidereal time caught up with NGC 3245; I then stopped my azimuth movement and started the Sprint.

NGC 3245 was one of the objects I observed in February with the 20-inch. My description then was: “205× Long bright nucleus running N-S, small diffuse halo, 2.3′ x 1′ visible. Easily found and bright.” Now with the 12.5-inch it was: “148× elliptical, glowing nucleus with DV, bright core with AV; 3:1 N-S. In a triangle of stars. Somewhat difficult to find.” I saw less of the galaxy, and it was dimmer. The Sky Quality Meter w/Lens (SQM-L) reading in February was 20.71; now it was 20.60 or so — about the same. This was to be one of the brightest objects I was to observe during the Sprint.

I abandoned the thought of doing a pure sweep. Because I had to navigate, and because I wanted some way of following the list, I decided to simply move the scope to the altitude of the next object on the list, and monitor the clock to wait for it to pass into view.

My clock didn’t display seconds so it was difficult to judge time precisely. Often the objects would be at nearly the same RA time but at different declinations, and I would spend too much time on one and miss the next. But the main problem was the difficulty in seeing the objects to begin with — I didn’t know how the object would appear, so I was scanning and using averted vision just to see if anything was in view. I sometimes suspected my calibration was slightly off, and I was like a ship whose bearing was wrong by a slight angle, causing it sail further off course over time. But from time to time, I would find an object at the listed location, and it would restore my confidence in my setup.

Certainly Herschel’s sky was not affected by light pollution, so he may have been able to see more at comparable apertures — we use larger apertures and higher altitudes to compensate for our light pollution. But I think my difficulty lay with my eye, which hasn’t been fully trained to see dim objects. I haven’t been at this long enough.

In the end I counted 28 “did not find” objects. 15 I regard as tentative – where my verbal description was not quite confirmed by my later check of the NGC/IC Project or Aladin websites, which I use post-session to confirm my observations. 30 objects I can confidently confirm.

My descriptions are composed primarily of the adjectives “very faint,” “exceedingly dim,” and “difficult.” But I don’t want to leave the wrong impression: I was really enjoying myself. Any night out under the stars is a good night, and here I was stretching my abilities and having fun.

There were some very good moments. My most exciting sight was NGC 4169, 4173, 4174, and 4175, a U-shaped cluster of galaxies known as The Box. It was incredible to see them drift across the view, one galaxy revealing itself after the next. I will admit, though, my biggest disappointment was the Coma Cluster. I had anticipated I would see a flurry of galaxies fly past my little window on the universe, and I was really looking forward to it. As it turned out, I missed the whole thing because I had missed one of the leading galaxies, which put my timing off for the rest. Since in this game there was no going back, I had to let them go.

The total Sprint takes about five and a half hours. There are gaps of 10, 15, or 20 minutes here and there where Herschel did not find any new object. He most likely kept working, but I took breaks.

Finally, after nearly three and a half hours, just after yet another exceedingly dim glow in NGC 5263, the great globular cluster M3 (NGC 5272) drifted into the eyepiece from the east in all its glory. I checked my inclinometer and my clock, and was very surprised to find I was only 0.2° off in my declination and ~30-60 seconds off in RA — little enough that M3 still floated cleanly into the FOV. It proved to me that my navigation was accurate, and the problem was the challenge of the objects and my own abilities.

M3 is not one of Herschel’s discoveries, but I am sure he would have let it pass through his eyepiece, and enjoy, as I did, feasting on a big, bright object after such labor. Full circle: M3 was the first globular I ever saw, from Monte Bello Open Space Preserve at the end of last April, found with the help of a kind fellow observer.

There is a long gap from M3 to the next object on the list. On the Uranometria charts there are a few NGC objects, but I see from the NGC/IC Project website they were discovered by others. There were also several Abell galaxy clusters, too dim for even Herschel to find. It was past 3am and I was tired, so I skipped forward and star-hopped to the last two galaxies.

I trained my scope on the Corona Cluster, just to see how it would look at this aperture and decide whether Herschel had a chance to see it. I detected some grayish mottling, but I saw no distinct object; this is probably what Herschel saw too, and could not recognize it for what it was. But all those dim ovals he could see, not clouds but with some structure, must have given him an inkling these were not simply “nebulae,” these were a different class of object altogether. If only he knew what we know now!

After so much dimness I craved a big bright galaxy, so I viewed M51. I could see both galaxies, and the extension between them, and with AV could trace spiral arms coiling in and out of the bright center. That was very satisfying.

I ended the night with a bang: Nova Sagittarii 2015 No. 2, which was just coming out over the hills at ~4:30am, found with the aid of the AAVSO chart. I was pretty worn out and didn’t try a magnitude comparison, but it was the brightest star in my binocular field. I later found its current estimate is magnitude 5.2.

I want to try the Sprint again, but next time with the 20-inch. I’ll find a way to restrain that scope’s altitude movement within the 3° band, so that I can leave the list behind and sweep as Herschel did, and enjoy the view to the fullest. On the bright side, with so many objects left unobserved on the list, it will still be a night of discovery for me.

— Mark McCarthy, Fremont, California, March 18, 2015


Image of

April 14, 2015 at 4:56 pm

Thanks for posting this... I’m impressed with your determination to follow closely in Herschel’s footsteps, especially considering your limited observing experience. This is by no means an “easy” observing project... best of luck to you on your next attempt!

You must be logged in to post a comment.

You must be logged in to post a comment.