Detecting the Interstellar Hydrogen Line, 1951
I prepared this text and the related slides for the Interstellar Hydrogen Workshop, held at NRAO in Green Bank WV in May 2001. At the Workshop I presented the chronology of events during 1950 and early 1951 that led to the "discovery" of the hydrogen line. Most of my time during that year was spent at the Harvard Nuclear Research Lab, just down the street from the Lyman Lab where I was assembling the hydrogen radio telescope on the weekends.
The Cyclotron page describes my work on the development of the 95" Harvard Cyclotron. A major highlight in Harvard Physics in 1951 was achievement of the first external beam. For two months I rarely left the Cyclotron Lab as I directed both a day and night shift to complete the installation of my external beam machine. On Feb 28, 1951, Harvard had its first external beam.
I then returned to the Lyman Lab for the first time since early December 1950. Major modifications to the receiver design were completed during the Thanksgiving holiday period in 1950, and then the entire hydrogen project was moth-balled pending completion of my cyclotron project. The modified receiver was turned on for the first time during the Easter weekend in 1951. By 3:00 AM on Sunday morning March 25, I was convinced that the line had been detected. The critical feature of the identification was the explanation of an unanticipated shift in the received frequency relative to the rest frequency. The possibility of a doppler shift was an almost immediate thought. I called the Harvard Observatory during the early hours of the morning, requesting details concerning velocity vector values associated with all motions between my telescope and a direction in space at a declination angle of -5° that passes through upper culmination at around 2 am. In answer to a question why I needed the information, I said that I was attempting to detect a hyperfine transition in interstellar hydrogen at a wavelength of 21cm and needed the velocity information to calculate the doppler shift. There was a moment of silence and then the telltale click as I was disconnected. I recall thinking at the time that the response merely indicated that I had been identified as one of those kooks that typically calls the Observatory during the wee hours to ask some stupid question.
So, without benefit of pocket calculator, I proceeded with the calculation based on my recollection of the values of the major motions from my days as an instructor in Astronomy at Amherst College in 1942. The results of my calculation identified the source of interstellar radiation as the hyperfine line of hydrogen. Within a week, the sidereal motion of the source was easily detected. I discussed the results with Ed Purcell and it was then that I learned that van deHulst was teaching at the Observatory. Ed suggested that I discuss the results with van deHulst. I recall asking Ed for a copy of the van deHulst paper in which he says that detection is highly unlikely. I was concerned that he might have forgotten the subject and the paper. That was definitely not the case. In fact, it was during this first meeting with van deHulst, while talking on the phone with Oort, that I learned that Leiden had been working on the detection of the 21cm line for several years. When Oort explained their approach, it was obvious why they had not succeeded. He put Mueller on the phone and I described the switch frequency technique that makes detection rather simple. I sent Mueller schematic drawings of the critical units, and also sent him some critical components, such as the acorn tube used in the IF preamplifier. Mueller was a very capable engineer and was able to detect the line in just a few weeks. I provided the same information to the Australian group and they too were able to detect the line in a few weeks. The Australians were not actively engaged in a search for the line, hence, their effort was a start-up program.
The second time that I met Van deHulst was at my doctorate oral. Ed Purcell had arranged that the three of us meet for this event. Following a brief introductory discussion of the equipment and its calibration we turned our attention to the physical processes involved in generating an emission line ~100 K above the background radiation field. The three of us worked on this subject for about 2 hours. We were unable to identify any mechanism that would explain the emission temperature of the line. In fact, we all agreed that Van deHulst’s negative prediction was valid, as presented in his publication of 1945 where he stated, "Discrete lines of hydrogen are proved to escape observation. The 2.12cm line, due to transitions between hyperfine structure components of the hydrogen ground level, might be observable if the life time of the upper level does not exceed 4.108 years, which however, is improbable."
Several months after the detection of the line, Ramsey and Weisskoff discovered a mechanism that would support the detection. None of the three papers published in Nature in 1951 provided the correct explanation of why the line was detectable.
Purcell and I could not resist asking Van deHulst why the Dutch Group had spent so many years trying to detect the line, when Van deHulst had proven it was undetectable. His answer was totally unanticipated. Apparently Oort was determined that the line would be detectable. He offered no logical reason other than the simple fact that a microwave line with an observable Doppler shift was needed in order to look through the dust clouds and develop a picture of the spiral nature of the galaxy. He was adamant. In Van deHulst’s words, "he willed the line to be there." Oort obtained the grants and other funding needed to pursue the detection. No one was allowed to express doubt that Oort could detect the line. At team meetings the discussion of detecting the line was very up beat, in fear that a negative comment might cost one his chance for a doctorate. Van deHulst admitted that he was no different than the others and would be up beat during the team meetings. A formal publication in a recognized technical journal was a different matter. Van deHulst felt it important to publish the results of his studies that showed the line was undetectable. To do otherwise might place his technical reputation in jeopardy.
The role of Oort in the detection of the line explains why I spent most of the time on the phone with Oort and very little time with van deHulst during our first meeting when I informed him of the line detection. It also explains the notes taken by a secretary during one of the team meetings quoting van deHulst as supporting the possible detection of the line. (I first learned about the secretary’s notes at the GBT meeting in 2001).
The panic mode of activity in December 1950 and January 1951 at the Cyclotron Lab was caused by a potential conflict between my external beam responsibility to the Lab and a request from the Navy to return to active duty for the Korean affair. I was unable to explain to the Navy that my acceptance to the Harvard Business School for the fall term of 1951 did not mean that I had completed my doctorate work in physics. I planned to obtain an MBA at the Business School during the two years it would take to write the "negative thesis" needed to obtain my doctorate in Physics. The Navy found this to be a convoluted argument and ordered me to report on May 1 to Washington DC. The February success with the external beam, followed by the detection of the line in March, completely upset the plan. There was no longer any reason why I should not report to the Navy on May 1, with my PhD in physics. There was no reason to pursue a degree at the Business School. I was on my way to Washington DC for 30 days to obtain a Q Clearance (nuclear security), then off to Los Alamos to join the bomb squad.
While in Washington awaiting my Q Clearance, I spent most of my time at NRL to find out what was new and interesting. During that time I learned about the considerable interest in underwater warfare and the beginning of the Polaris Program, which was initially to be installed on large ships the size of a cruiser. The submarine approach, however, was on the drawing board. I proposed that we look into the application of radio astronomy sensor techniques to the underwater detection problem. I was awarded a 50K contract for the study. That was the end of my Korean service.
I returned to Cambridge and introduced myself and my contract to the president of the Scientific Specialties Corporation. The president, Tex Holt, was the Director of the Nuclear Lab at Harvard at the time that I obtained the external beam. Bell Lab announced the discovery of the transistor in 1951. Tex bought a license from Bell to manufacture transistors. My underwater work was viewed as a Special Project, outside of the main business stream of Scientific Specialties. Consequently, I was able to operate my project without direction or interference. I rented a 65’ yacht with crew for the summer from another good friend at the Nuclear Lab, and spent most of the summer of 1951 cruising the waters of Buzzard’s Bay and the Islands. In the process, my team moved underwater detection from an ear phone approach to a machine approach capable of sorting man-made noises from natural background noises.
The comfortable life of an audio engineer on a yacht in Boston Harbor was about to come to an end, however, with the return of Bart Bok from South Africa. That part of the story is continued in the description of the Harvard 24ft and 60ft.
Editor's note: On this page and in the related slides Ewen describes the equipment, the detection of the HI line, and events surrounding that detection. For a more detailed account of the methodology and results, see the initial detection papers, in a 1951 Nature section titled "Observation of a Line in the Galactic Radio Spectrum," as well as the 1999 description of the detection:
- Ewen, H.I. and Purcell, E.M. 1951. "Radiation from Galactic Hydrogen at 1420 Mc./sec" Nature 168: 356, 1951.
- Muller, C.A. and Oort, J.H. 1951. "The Interstellar Hydrogen Line at 1420 Mc./sec, and an Estimate of Galactic Rotation" Nature 168: 357-358.
- See also the note by J.L. Pawsey following the paper by Muller and Oort. The Australian results were published more fully in 1952 by Christiansen and Hindman: Australian Journal of Scientific Research A5, 437-455.
- Stephan, K.D. 1999. "How Ewen and Purcell Discovered the 21 cm Interstellar Hydrogen Line" IEEE Antennas and Propagation Magazine 41 (1): 7-17.
View related HI slides
View slides of 2001 visit to Green Bank