Interview with Alan H. Barrett, 1971

Sullivan: 00:02

This is August '71 at the Amherst AAS meeting interview with Al Barrett. So what was the story on this OH search at NRL? Was it your impetus that got it started?

Barrett: 00:21

Well, it's hard to say whose impetus. When I got [inaudible] in exchange of correspondence between Ed Lilley and George Dousmanis. George Dousmanis and I were office mates at Columbia. And he was doing his PhD about the same time I was. But his PhD was on the microwave spectra of OH, not at 18 centimeters, but at higher frequencies. But because Townes had made him aware of the possibilities of OH for interstellar radio astronomy, Dousmanis had calculated where the 18 centimeter lines would fall, so. And because Townes was my thesis advisor, we had talked about OH. So it was understood, and in fact, was stated in my application for employment at NRL, or application for a postdoctoral fellowship, that a part of my time would be spent doing an OH search. Because Townes was my thesis advisor, my office mate was George Dousmanis, it's just an obvious thing to go ahead with.

Sullivan: 01:56

Was it measured in the lab or only calculated theoretically?

Barrett: 02:00

No, it wasn't measured in the lab at that time.

Sullivan: 02:04

Yeah, when you did that search.

Barrett: 02:05

Yeah. What Dousmanis had done was measure a number of microwave states of OH. But his apparatus was such that it wouldn't propagate 18 centimeters. So he didn't measure 18 centimeters. And in fact, I don't think he measured anything less than-- well, I'd have to check in the literature here, but I think it's about 15 gigahertz or about 2 centimeters and above in measure. But once you know the constants of the molecule, you can calculate down to what it ought to be 18 centimeters if the theory is in good shape. Well, the only theory of lambda of doubling that existed then was Van Vleck’s, which was done in something like 1927. So Dousmanis expanded the theory and added higher order terms than the Hamiltonian and took these into account. And because of his measurements then, he knew the molecular constants and he figured he could work back and arrive at a reasonable number for 18 centimeters. And as I recall, the uncertainty there was about 10 megacycles. So that's how we got into doing that.

Barrett: 03:36

Now, we started that search-- Let's see, I got there in January.

Sullivan: 03:47

'56?

Barrett: 03:48

In '56.

Sullivan: 03:50

This is on the 50-foot?

Barrett: 03:52

Yeah. It was the 50-foot dish on the roof. And the first thing I had to do was build a horn because we had no means of illuminating it. I built the horn and I tested the horn out. And I guess a mixer was also built because all they had at that time was a 21-centimeter receiver. And then we went ahead with the search from there. I guess it started sometime in the summertime. I know I recall arriving at NRL in the evening for the evening shift when it was still daylight. So it must have been sometime in summertime.

Sullivan: 04:31

Now was this an extensive sort of search? Many sources, reasonable with integration time or did you-- ?

Barrett: 04:42

Well, we were using the hydrogen line observations as a guide as to where to look, and most of our time, not all of it, but most of it, was spent on Cassiopeia, because it was circumpolar and because it had the sharp hydrogen line of [inaudible] features. But we also looked occasionally, I guess, at Taurus and the Galactic Center.

Sullivan: 05:09

Did you search over this whole I guess it would be [inaudible]?

Barrett: 05:12

Oh, yes, all the time we were on any one source, we were also scanning in frequency. And it was obvious then, obvious in quotation marks, that we should be searching at 1667. We now know that that's not entirely the case. But that's what we did. Now, we had a bit of an alarm at one point because we saw something, but we saw it in emission and we didn't understand it. It was a fairly strong signal then.

Sullivan: 05:54

In Cas was this?

Barrett: 05:55

By standards of those days. Yes, it was in Cas. And I don't know, it was maybe 20, 30, 40 degrees, I don't recall, but it was strong. But it was in emission, and that bothered us. And we chased that around quite a bit, and we finally figured out that we were observing a meteorological balloon in our other image. I don't know if it's upper or lower, that doesn't matter, but in the other image of our pass band. Or wait a minute, I'm not entirely certain about that. We were frequency switching and maybe it was just in the other band of switching.

Sullivan: 06:44

Yeah, either way.

Barrett: 06:46

But that would have put it in obstruction, wouldn't it? Anyway.

Sullivan: 06:50

Yeah, that would have [crosstalk] transmitter.

Barrett: 06:52

I think so. Yeah, I think so. Well, it was in the other image. But we went off Cas, I recall the excitement about that, we went off Cas and it didn't go away, so.

Sullivan: 07:08

[inaudible] rule it out.

Barrett: 07:08

Yes, but we were stuck. But we did chase that around.

Sullivan: 07:13

Well, in retrospect, should you have seen the OH at that time? What do you think is the story? First of all, that estimated frequency, does that cover the correct frequency?

Barrett: 07:30

Yeah. We were scanning the proper frequency range, and in retrospect, knowing now what we know about gas absorption, no, we would not have seen it.

Sullivan: 07:45

Your sensitivity was not enough.

Barrett: 07:46

That's right, that's right. The electronics in 1956 wasn't what it was in 1963, and it was a smaller antenna, and we just never would have seen it. So we undoubtedly scanned the right range, but we never saw it. Now, had we looked at some of these unusual sources where OH was amplifying, maybe we would have seen it.

Sullivan: 08:09

But of course, there was nothing particular about those sources at that time. In fact, the Westerhout catalog wasn't even existence then.

Barrett: 08:14

Oh, no. We had absolutely no reason at all to imagine anything like an amplifier, in intersteller--

Sullivan: 08:21

Or even to know to look at something like W3 or W49, or anything like that.

Barrett: 08:25

That's right. That's right. Subsequent events, if we had pointed the antenna at the right place, scanned the right frequency, we would see it. But that's all hindsight.

Sullivan: 08:43

Yeah. Now, what about the eventual discovery? I guess it was in '63, apparently you continued your interest in looking for OH.

Barrett: 08:53

Oh, yes. Oh, yes.

Sullivan: 08:55

Was this the next time you tried?

Barrett: 08:57

Sullivan:

Barrett:

Yep, yep. I was in Michigan and there was no possibility there. They were just beginning their program in radio astronomy and there was no possibility of any OH studies. But in 1961, I went to MIT and one of the programs uppermost in my mind then was to have another look at OH because in the intervening five years, Townes and his students had measured the ground state lambda doubling of OH. So that automatically removed one of the three variables that was involved in the search, alpha and delta and frequency. So we knew now what the frequency was and nobody else had bothered to look. So that was one of the things that uppermost on my mind. Now, along with this was the fact that Jerry Wiesner had a graduate student named Sandy Weinreb who was working on a new spectral line device, the autocorrelator. And it was a natural marriage between my interest in OH, Sandy Weinreb’s device, autocorrelator, to marry the two of them and look for OH.

The only other problem, I guess, is to find an antenna.

Well, Sandy's thesis was to build the correlator and use it on an astronomical measurement. And this he did. He used two astronomical measurements, both of them negative. One was the Zeeman splitting in Cassiopeia, and I guess Taurus, I'm not certain about that, I have to look at the literature. And the other was the 327 megacycle, deuterium switch. As I said, they were both negative. But he got his PhD. And he went to work for Lincoln Laboratories.

Barrett: 11:34

Sullivan:

Barrett:

And then I arrived at Tech to start up radio astronomy. There was no radio astronomy at MIT before I got there. And the question was, what antenna do you use? Haystack wasn't in existence.

Did you go to MIT especially to start a radio astronomy program?

Oh, yes. Well, there was a Millstone Antenna, 84 feet, at Lincoln Laboratories which was used off and on for various military and other programs, and we began to explore the possibility of using that antenna for radio astronomy. At that time, Lit Meeks was also at Lincoln Laboratories, and both Lit and Sandy were responsive--

Sullivan: 12:29

This is an interview.

Barrett: 12:31

Hi, Don. How are you? That's all right. Be recorded. Don't be bashful.

Sullivan: 12:36

Say Hi.

Unknown: 12:37

Hi.

Barrett: 12:41

At that time, both Lit and Sandy were responsive to using the correlator on OH [problems?]. So that was the beginning, I guess, if there has to be a beginning, of that experiment. And it began in, I suppose, maybe late '62 or early '63. By October of '63, the receiver was on the antenna, the correlator was built, and a fellow by the name of Henry-- John Henry, I think, but you better check this.

Sullivan: 13:25

J.C. Henry, I think.

Barrett: 13:26

Yeah. I don't know. Had written the program, for doing the transform on the correlator, into the computer that they had at Millstone at that time. It had no commercial number because it was one that was built in MIT. And he was one of the ones who, if he wasn't instrumental in building it, at least he was the only one we could rely upon to actually tie the thing in to give us an answer.

Sullivan: 14:02

You had an online display then?

Barrett: 14:03

We had an online display, finally, but only because of Henry's efforts. Well, the experiment began in October, with the correlator, with a mixer radiometer in the Millstone antenna, and with Henry's programs to give us the spectrum out of the back end. And the first couple of nights - at least the first night, and I think maybe a couple of nights - the results were negative. And we had standing wave problems and mismatches in the antenna and in the receiver inputs. But eventually, these were solved. And the first real good night, where we were hopeful of actually seeing something, we saw something, which was again on Cassiopeia, and there was an indication of a line within the first half hour. Well, that makes a great statement, but the first half hour of observing, when you see something, you don't believe it. You go back and check it and you do all kinds of things to make it go away. And in this case, it wouldn't go away and we got very excited about it. And we didn't have the antenna every night, but every time we did, we would show up and we'd do different things, and the line would still be there.

Sullivan: 15:31

Was this the first astronomy that was ever done on that antenna, or had it done some radar astronomy, maybe earlier? [inaudible]

Barrett: 15:39

I don't really have an answer on that. It was the first radio astronomy done, but that wasn't your question.

Sullivan: 15:44

Right. The radar.

Barrett: 15:46

The radar astronomy. This was the antenna that initially recorded seeing Venus, and had a value of the AU as a result of having seen Venus, which ultimately was an error. So I'm hesitant to answer that question. It did some radar astronomy, but it was wrong. So there you are.

Sullivan: 16:10

So you found it in Cas-- ?

Barrett: 16:11

Yes.

Sullivan: 16:12

And you found both lines, right? So you were pretty sure of the identification, I guess?

Barrett: 16:18

Mm-hmm. We found the first line. The 1667 line. It was a couple of degrees, as I recall. And we did the usual test of moving the line and seeing if it would move and going off source and seeing if it would go away, which it did. We looked for 1665, which was smaller. It was of the order of one degree and it was there. But the only thing we had to go on in this whole thing was the hydrogen line detections, at that point, and we went back and looked up how they verified they had it. And you wait a while until the motion of the earth around the sun gives you the proper Doppler shift.

Sullivan: 17:11

Different Doppler shifts, yeah.

Barrett: 17:13

And I tell you, those were--

Sullivan: 17:15

Now, what do you mean by wait a while? During a day?

Barrett: 17:17

No, it was two weeks.

Sullivan: 17:18

Two weeks.

Barrett: 17:19

It was two weeks. In two weeks' time, the thing shifted by ten kilocycles, which are now called kilohertz. But we were very, very nip and tuck. Very touchy and excited and concerned over that two-week time span to see if it would move

Sullivan: 17:39

You didn't have enough resolution, I guess, with the Earth's rotation to be able to pick that difference up. That's a half a kilometer of seconds. I guess that wasn't quite enough.

Barrett: 17:49

No, we weren't working with that resolution. And we didn't have the dish that long. And we had it mostly at night when nobody else wanted it. We did look in the Galactic Center, and we had horrible baseline problems, which we now know, as the Australians found out that. But the baseline problems were the fact that--

Sullivan: 18:10

It was a huge [crosstalk].

Barrett: 18:12

That it was a very small absorption feature on the edge of a much larger absorption feature because we were using the hydrogen line observations as a guide as they did. And if you look at the first Australian paper on OH, which was the first formation of OH, there's no spectra given at all. That's because they didn't understand the baseline problem. And in fact, it was a small dip on the edge of a very, very large dip.

Sullivan: 18:39

Was this in Nature, this Australian paper you're thinking?

Barrett: 18:43

The Australian paper was in Nature. We published our paper in Nature. And there's a peculiar aspect of this, which I suppose ought to be recorded because we discovered this right at the time that the ITTU or ITU [International Telecommunications Union] or something or other was convening in Switzerland to allocate frequencies. Well, we had submitted our paper to Nature. And we were worried about if we told the American representative of the ITU in Switzerland that we had seen this, and if the press picked it up, would Nature publish our paper? So what should we do? Well, the press office at MIT, the public relations office at MIT was very helpful in this respect. We told them our problem, and it was agreed we ought to send the American radio astronomer accompanying the American delegation to the ITU telegraph, which would tell them about the detection, name the frequencies, because at that time they weren't all that tied down. And write the editor of Nature that we were doing this, and hoping that no matter what happened we wouldn't prejudice our case for publication in Nature. Well, this we did. And we sent George Swenson of the University of Illinois, who was the radio astronomer accompanying the American delegation, that we had seen these things. And we gave them the frequencies.

Barrett: 20:39

What we didn't know was he posted it on the bulletin board at the ITU. And he called van de Hulst at Leiden. And he asked van de Hulst, "What is the possibility that this is atmospherical age?" And we had given the frequency to about-- I think it was seven significant figures. Van de Hulst realized that if we knew it that well, it had to be interstellar. Because if it was atmospheric, it would be broadened more than it occurred. But also, Australia had representatives at the ITU. And the Australian representative picked this thing up off the bulletin board, wrote down the frequencies, and cabled Bolton in Australia that OH had been found at these frequencies in interstellar space. Well, as a result, the Australians had a little bit of a head start on the confirmation ahead of anybody else because they had the frequencies. Well, after we had sent Swenson the telegram, we began to have second thoughts about the advisability of doing this, you see. So I called him from my in home one night, in Switzerland, and we explained the concern. He told me he had spoken to van de Hulst. We explained all the details of the experiment and confirmed to his agreement or his satisfaction that it was indeed interstellar OH we were seeing. The following morning, he took down the telegram that had been posted, but the damage had been done to a certain extent. The press hadn't picked it up, which was good, but the Australians had.

Sullivan: 22:37

And so they very soon, I guess, got a [inaudible] receiver going on their own?

Barrett: 22:41

Well, the Australians moved on it very quickly and in no time at all, and I don't know what no time here means, I could go back and dig the dates out, perhaps. But between the time we had sent Nature our letter and the time it was published, I got a cable from the Australian Embassy in Washington that had received a cable from John Bolton in Australia. They had confirmed the existence of OH in the Galactic Center.

Sullivan: 23:13

You didn't even know that they were doing it at all, or you didn't even know that they knew.

Barrett: 23:16

No, I hadn't then.

Sullivan: 23:19

So you wondered how the hell they could confirm it since it wasn't even known at all?

Barrett: 23:23

No, I did know. That's right. I did know. And the reason I knew was because the Australian representative at the ITU Geneva had talked to George Wilson. So he told me that the Australians already had it, so I knew that they were onto it.

Sullivan: 23:43

Yeah. Was this a radio astronomer or the Australian representative? Or is it--?

Barrett: 23:48

His name I don't recall, but his name I was told at the time it wasn't anybody that actually rang a bell. So I think he was an administrative type or an engineer type. Well, I immediately responded to the cable by Bolton in that I was thinking again of the hydrogen line situation, where there were two of the three papers of the hydrogen line thing occurring in the same issue of Nature. I sent him back a cable saying that our paper would be published thus and so issue of Nature, is there a possibility he could get a cable in confirming it? Well, I didn't get an answer by cable. Eventually, I got an answer by a letter that no, that they had already missed that deadline and they weren't all that certain of their data and the reason why they were having these baseline problems, you see. So that whole thing went by the boards and our paper finally appeared, and eventually, there was a series of publications from Australia on the Galactic Center. But the first one of which had no spectra whatsoever.

Sullivan: 25:02

Now, did you follow this work up at all at that time, either in absorption or looking at other sources, or?

Barrett: 25:10

Yeah, we kept at it mainly in Cassiopea. There was a problem with observing time, and we didn't have too much observing time. And then I went off to Russia, and when we came back, Haystack was in operation. And then we went into operation in Haystack on a full bore basis.

Sullivan: 25:43

And that can be--

Barrett: 25:45

There, we found three things that were important. The first thing we found was the Cassiopeia line. If you look at it, the high resolution was split. Now I've got to go back and check whether that was Haystack or Millstone. I think it was Haystack. It was split. It was a single line at the hydrogen frequencies. It was a double line in high resolution at OH. We also found then, that at that time, this was now 1964, almost a year later, the so-called mysterium began to pop up. At OH we found an emission. So we started studying that, and we immediately found, which others hadn't done yet because they hadn't gotten around to it, it was polarized. And then we very rapidly found it was circularly polarized.

Sullivan: 26:41

This was the paper of science and W3(OH) that came out of that?

Barrett: 26:44

There were two of them, one of which that it was linear polarized and then there was a follow on paper that said it was found to be circular. And by that time, Al Rogers was involved in the OH studies too.

Sullivan: 26:59

On the OH absorption, was there any other group at the time working on that the same time--

Barrett: 27:04

Absorption?

Sullivan: 27:05

Yeah. Initially working on that more or less in competition with you, or--?

Barrett: 27:10

Oh, yeah, lots of people jumped on it right away.

Sullivan: 27:12

No. No. I mean, not after they knew about the discovery, but simultaneously, or--

Barrett: 27:20

Oh yeah. Yeah. Yeah. Yeah.

Sullivan: 27:22

--or even before.

Barrett: 27:23

Yeah. I went to a meeting of people who were concerned with the radio properties of Venus. This meeting was held out in Pasadena, California, JPL. And one of the speakers at that meeting was a radio astronomer from Pasadena. And this was early '63, when we hadn't yet discovered OH. And I think it was Barry Clark, but I'm not absolutely certain about that. I'm reasonably certain. But he gave a paper about the radio properties of Venus at L band. And the peculiar thing about it was that his L band frequency was 1666, not 1600, or even 1660. It was 1666. And I knew at that time we were instrumenting Millstone to go ahead with OH at 1665 and 1667. So obviously, I got concerned. Why did he have a frequency down at 1666 if he wasn't really concerned with this frequency? So I figured he was using that equipment for OH. And we were very concerned about that at the time we were doing our search. And after we knew we had it and we're trying to publish it, we were worried that we'd get scooped.

Barrett: 28:58

Well, I was talking later with Radhakrishnan, and he was the one who was looking for OH, and he was indeed looking for OH at Caltech. And I don't know how much off the record this one ought to be, but he told me he was getting no support for his interest - and I hope I get to edit some of this and look it over - but he was getting no support for his interest in OH at Caltech. But he did make one look one night when the equipment didn't work well. And his results, of course, were negative under those circumstances. It was his first look. And if you recall, I told you earlier we had a number of nights when we didn't have things tuned up and we just had negative results too, when you make the usual mistakes you make the first time. But he did indeed look, and I don't know the actual dates but I would guess that he looked ahead of us, but had equipment problems and never went back to it, didn't push it, and didn't see it.

Sullivan: 30:09

Was there any other group that you knew about?

Barrett: 30:12

No, I don't think there was any other group. I did have a letter when I was still at the University of Michigan. This must have been back in the late '50s. I had a letter from somebody at Jodrell Bank while I was at Michigan asking me about OH. No, it wasn't the late '50s, it must have been 1960 because I wrote back that I was leaving Michigan, going to Tech. One of the things I was very interested in doing, at the top of my list, was the OH search again.

Sullivan: 30:46

That was their interest also, I take it.

Barrett: 30:48

What?

Sullivan: 30:48

That was their interest also, I take it.

Barrett: 30:50

I guess. But I asked Lovell about this a number of years later and he had no idea that the whole thing was underway. So I don't think it was underway. I think I heard from a graduate student who was just asking a question. So I don't honestly think they were very serious about it, because after all, he would have known about it had they been.

Sullivan: 31:15

Now what about the emission that I guess was discovered at Berkeley, right, when they called it mysterium and so forth.

Barrett: 31:28

I don’t want to get drawn into the argument as to where it was discovered. Ellen Gundermann at Harvard, and Harold Weaver and his graduate students at Berkeley, as far as I know, discovered it essentially the same time. But the Berkeley group was the first one to print.

Sullivan: 31:43

Right. But how did it escape your attention? Or is it just a matter of telescope time, you didn't have any more, or you weren't thinking of those sources? Because even then there was nothing peculiar about W3 and W49 typically, even though the Westerhout Catalog was out, in 1963, 64.

Barrett: 32:08

it escaped our attention for a number of reasons. One was telescope time. Another was it was at 1665 and not 1667, where it was strongest. And I spent a large part-- not a large part, I spent three months of 1963, the last part of 1963, in Russia.

Sullivan: 32:35

By the time you got back, it had been discovered.

Barrett: 32:39

No, it was discovered in early '64. But you see, there wasn't that much time for radio astronomy on Millstone, and Haystack wasn't yet operational. So as I recall, the first thing we did on Haystack at OH was either we looked at Cassiopeia with high resolution, we had a bigger signal then, or maybe that was done at Millstone. I'll have to check that [inaudible]. But as I recall, the first real excitement at Haystack was on the emission and the fact that it was polarized.

Sullivan: 33:21

Yeah. Well, Ed Meeks tells me that there is an archive at the Millstone library, or that some secretary has one there, so I'll definitely want to be taking a look at that on the OH discovery.

Barrett: 33:33

I'm sure there is.

Sullivan: 33:37

Well, that pretty well--

Barrett: 33:38

I hope we're pretty near through because I got to go.

Sullivan: 33:39

That pretty well covers that one. I was just wondering if there's any other thing in radio astronomy of note, that you think you might be able to contribute to. I may not know the history as well as you do, as things you were involved in and so forth.

Barrett: 33:56

Well, I don't know. Not right now, maybe sometime we can chat about the first radio astronomy experiment done from a space vehicle.

Sullivan: 34:05

Which was that?

Barrett: 34:07

The microwave measurements of Venus. [crosstalk].

Sullivan: 34:09

Oh, oh, that's right. I forgotten that was-- yeah.

Barrett: 34:12

But I don't want to get you started on that now and I don't want to get started on that.

Sullivan: 34:15

Well, yeah, if you could just tell me anything else though. Like I said, you don't necessarily have to go into it now.

Barrett: 34:21

Anything besides that?

Sullivan: 34:21

Yeah.

Barrett: 34:26

Oh, I suppose, Woody, if I give it some thought [inaudible]. I was at NRL when the first Venus measurements were made, the first planetary measurements. Well, I don't mean the first planetary measurements, because obviously, they were Jupiter. But I mean the first microwave thermal emissions.

Sullivan: 34:49

Was there any line searching in Michigan or anything of that nature?

Barrett: 34:53

None. No. Not while I was there.

Sullivan: 34:56

Yeah. Okay. Well, thank you very much. I appreciate your time.

Barrett: 35:01

Okay. Maybe we can talk about Venus some other time.

Sullivan: 35:03

Yeah, sure.

Barrett: 35:05

But I want to get to New Hampshire now.

Sullivan: 35:07

I can understand. I've just been--

Barrett: 35:09

You thought I want to get to eat.

Sullivan: 35:10

That ends the interview with Al Barrett at the AAS in Amherst, August 1971.