The key to Michelson’s work was the Mount Wilson station. It’s where everything happened and the measurements were made. In essence, a light pulse was sent off to a distant mountain top where a mirror reflected it back to Mount Wilson and its travel time measured. Sounds simple, but to actually pull it off in the 1920s, with Steampunk style equipment….well that was just something beyond elegant.
Seeing as how this was well before lasers were ever imagined, sending a light beam over long distances without it expanding to a useless diameter was a challenge. Michelson accomplished this by using a Sperry arc lamp, one of the most powerful of the day. If the operators weren’t careful, they’d be burned by its ultraviolet emissions. The arc light was focused into a somewhat parallel beam by 24″ diameter concave mirrors at each end of the link (Later experiments used 40″ diameter mirrors). Given the other necessary optics involved, aligning the system and keeping it that way was difficult.
The actual measurement process was exceptionally clever. The primary measurement point, what Michelson called “Mount San Antonio” (today known as Lookout Mountain) was 22 miles away to the northeast. The light pulses sent from Mount Wilson to San Antonio were created by reflecting the arc light off a spinning, multifaceted mirror with either 8, 12 or 16 sides. The time it took a pulse of light to travel from Mount Wilson to Mount San Antonio and back to Mount Wilson (44 miles) was 0.00023 seconds. This happened to be the same amount of time it took a 16 faceted mirror, turning at 264 revolutions per second, to move one facet. When all was adjusted just so, the light pulse went out on one mirror facet and returned on the next one presenting what appeared to be a single image to the observer.
When using mirrors with smaller numbers of facets, the rotational speed had to be increased accordingly. This created some dangerous conditions. At one point Michelson attempted to use a larger, 8 sided mirror to obtain a brighter return image. It was designed to operate at 528 turns per second (almost 32,000 RPM). But when it reached 400 revolutions per second it burst apart, showering the observation room with shards of glass. And while it was possible to place a shield around the spinning mirror, the line between it and the observers eyepiece had to be kept clear. Not a job for the timid.
The spinning mirror was driven by compressed air, continually adjusted by the observer. Timing was performed by a rather convoluted process starting with a highly calibrated free pendulum, and feeding eventually into an electric tuning fork.
Because the mirrors had a 30 foot focal length, Michelson’s initial installation working with Lookout Mountain was a long structure, about 35 feet. In the later phases of his experimentation, when he attempted to use 40″ diameter mirrors to bridge the distance to Pine Cove (San Jacinto) then later Santiago Peak, Michelson constructed two separate huts, connected by a tunnel structure to protect the light beam.
Michelson used his Mount Wilson station from 1922 to 1928. However what Michelson called his “definitive measurements”, between Mount Wilson and Mount San Antonio, were made during the Summer of 1926. Michelson was originally brought to Mount Wilson by the facility’s first Director, George Ellery Hale. Hale was succeeded as Director in 1923 by Walter S. Adams, who worked closely with Michelson until Michelson’s death in 1931.
The site today
Michelson’s Mount Wilson station was located on a ridge on the southerly face of the mountain, in what is today a residential area for visiting astronomers. Due to the fact these folks need to sleep during the day to work at night, this area isn’t open to the public for visitation.
I’d like to offer up special thanks to Maggie Moran, Mount Wilson Site Superintendent, who generously allowed me access from December 2017 through January 2018 to survey the site in detail with a total station instrument. This resulted in a few surprises and some mysteries solved. Also additional thanks to Ken Evans and Bill Leflang, volunteer docents at Mount Wilson for assisting me in acquiring access and digging up a lot of excellent background info. And final thanks to Larry Webster, an employee at Mt. Wilson for the CHARRA array, who unearthed the actual engineering drawings Michelson and his staff used to construct the facilities and apparatus. Those were amazing!
Surveying the site in detail was a wonderful experience. Not only for figuring out what the concrete foundations and wood posts were originally for, but the amazing views to be had when visibility cooperated. Below is a zoomed picture I took of downtown Los Angeles from the Michelson site. Just above city center, near the coast, can be seen the runways of LAX.
The foundations of Michelson’s huts remain, along with a commemorative plaque and the benchmark of the US Coast and Geodetic Survey from 1922, labeled “Michelson”.
The Mount Wilson Station’s three phases
Mount Wilson to Lookout Mountain, Circa 1924 – 1926
As mentioned, this was Michelson’s first, and likely best, measurement effort, utilizing 24″ diameter mirrors at each end. Below is a drawing, based upon my site survey and filed plans, as to what that installation looked like. You can also download a pdf of the drawing, linked here , that allows for zooming all the way in to look at the beam splitter and mirror mounts, drawn to scale from Michelson’s plans.
There is a bit of uncertainty as to the precise layout of this initial structure of Michelson’s. While his later huts were well documented, this first one was not, or at least what documentation may have once existed is now gone. However based upon a few hand drawn notes, pier locations, mirror focal lengths and the San Jacinto plans showing the outline of the existing, “as-built” structure, I believe I have accurately depicted its layout.
A bit more ambiguity exists in regards to the shape and precise location of the arc light building. This was apparently a separate, but adjacent, structure which housed the noxious and nasty arc light source. Based upon the located anchor points for the survey tower, and the fact the light source needed to enter the splitter apparatus from its right side, I made an educated guess regarding its placement.
Note the survey tower to the right of the main building. This tower was a dual tower, centered over the benchmark established by the US Coast and Geodetic Survey measuring the distance to Lookout Mountain. The “dual” aspect meant it was a tower inside a tower. The survey instrument rested on the inside tower, physically separated from the outer tower which personnel worked on. This greatly reduced instrument vibration and resulted in greater accuracy. Fun fact: These later became known as “Bilby Towers” which saw extensive use in high accuracy surveys. Not coincidentally, the name of the survey signalman in the party was one J. S. Bilby. This was likely one of the very first installations of this tower style.
Mount Wilson to San Jacinto, Circa 1926 – 1927
When Michelson shifted his target to San Jacinto, in an attempt to measure over a longer distance for better accuracy, he made the jump to 40″ diameter mirrors at each end. These 40″ mirrors increased the light gathering by 2.8 times over his original 24″ mirrors. Unfortunately the path length of the light beam increased from 22 miles to 82 miles, a factor of 3.7 times. This, at least in part, probably contributed to the ultimate failure of utilizing San Jacinto as a target.
Of course the new mirrors required all new mounts, piers and huts. Fortunately their construction was well documented in the Mount Wilson archives, and backed up by measured locations during the site mapping. The site as configured for the San Jacinto attempt is shown below, and a high-res pdf of the image below may be grabbed here
Mount Wilson to Santiago Peak, Circa 1928
With the failure of the San Jacinto effort, Michelson switched to a closer target on Santiago Peak in what is today Orange County. He reused the same mirrors and mounts, constructed new concrete piers and simply rotated the Mount Wilson station approximately 27 degrees. A high-res pdf of the image below may be grabbed here .
Again, the 40″ diameter mirrors represented a 2.8 times increase in light gathering ability over the 24″ mirrors, and this time the path length of 46 miles to Santiago Peak represented a 2.1 times increase over the Lookout Mountain path length. So in theory, with the increase in light gathering being greater than the increase in path length, that link should have worked. But it did not.
At the Huntington Library I came across a photo of one of Michelson’s installations, of which I had seen other versions online. It was unclear just which of Michelson’s incarnations this photo was. Thanks to the Huntington, I was able to obtain a clean, high-res copy. Once my site survey was complete and accurate bearing information was obtained, it was possible to determine the photo was of the configuration utilized for the Santiago Peak work. With that known, I was able to do a “Now and Then” comparison of the site, which was rather interesting, and the Bilby-style survey tower is visible behind the left hut:
Back to Historic Speed of Light measurements in Southern California