Interview with Maarten Schmidt
Description
Creator
Rights
Type
Interviewer
Interviewee
Location
Original Format of Digital Item
Duration
Interview Date
Interview Topics
Notes
The interview listed below was either transcribed as part of Sullivan's research for his book, Cosmic Noise: A History of Early Radio Astronomy (Cambridge University Press, 2009) or was transcribed in the NRAO Archives by Sierra Smith in 2012-2013. The transcription may have been read and edited for clarity by Sullivan, and may have also been read and edited by the interviewee. Any notes added in the reading/editing process by Sullivan, the interviewee, or others who read the transcript have been included in brackets. If the interview was transcribed for Sullivan, the original typescript of the interview is available in the NRAO Archives. Sullivan's notes about each interview are available on the individual interviewee's Web page. During processing, full names of institutions and people were added in brackets and if especially long the interview was split into parts reflecting the sides of the original audio cassette tapes. We are grateful for the 2011 Herbert C. Pollock Award from Dudley Observatory which funded digitization of the original cassette tapes, and for a 2012 grant from American Institute of Physics, Center for the History of Physics, which funded the work of posting these interviews to the Web.
Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event.
Series
Unit
Transcription
Sullivan
This is interviewing Maarten Schmidt on 18 August ’75 at the AAS [American Astronomical Society] Meeting at San Diego. So when did you first come into contact with radio astronomy?
Schmidt
I think it must have been late 1953 or early 1954. After I had entered graduate school at Leiden in 1949, [Jan Hendrik] Oort, the Director, asked me to go to Kenya to do observations with [?] to determine declinations of stars from a mountain top in the crater. So my study got a bit delayed, because I only came back in December, 1951 and, in the meantime, of course, all the excitement of the discovery of the 21 cm line had happened with all three places having been involved with [Harold Irving "Doc"] Ewing at Harvard, and the Dutch and the Australians, all in a rather short time span, which I think was the middle of ‘51. I was completely out of that, I mean I was in Kenya in the woods. And when I came back all that had happened and I had to resume my graduate studies, so I'm sure I took classes for the next two years or so and toward the end of ‘53, I think, I got involved in radio astronomy, which was to be the second approximation, the second goal at spiral structure. The first one had been done in a paper, by I think, by Oort and [Hendrik C.] Van de Hulst and [Christiaan Alexander "Lex"] Muller or so when the external, that is to the Sun, the spiral structure had been investigated and uncovered for the first time. And this was to be a second go-around where I was to take the interior part of the Galaxy, within the solar distance, and of course, over the northern hemisphere, and [Gart] Westerhout was to take the external part of the Galaxy outside the solar distance.
Sullivan
The neutral hydrogen?
Schmidt
Neutral hydrogen, all at 21 cm.
Sullivan
And I assume that it was Professor Oort that was directing, choreographing all of this?
Schmidt
Yes, oh, yes. He had a grand plan that he didn't necessarily talk about that much, but we were part of that, of course, at that time already. And so certainly in 1954, I went up twice I'm sure in the beginning or later in that year for a fourteen day or three week run with a Kootwjik dish which had to be moved by hand every two and a half minutes both in azimuth and altitude, and the basis of pre-computed tables, I remember that one got very good at reading a book or just studying all about two minutes and thirty seconds, then to look up at the clock to confirm that, "Yes, it was about time to change things." And then to change it, you know, from inside the observing cabin...
Sullivan
And these predictions, did they come from, did you have a [?] that...
Schmidt
Yes, we had a [?], human computers, not the machine computer, but the human computers, had all prepared that. It was quite a lot of work, really. Anyhow, I remember one of the things I may have started was to, instead of doing line profiles, which would be at a particular position given a longitude and latitude and then as a function of frequency, that I started to do cross-sections across the Milky Way more or less in latitude at a given longitude more or less, and certainly a given radio velocity or frequency. Because the difficulty, of course, of the detection and disentangling of spiral structure in the interior part of the Galaxy is that there are two points along the line of sight at equal galactic centric distance where at least, in the plane the two contributions are, from the two points, cannot be separated out. It is only as a function of latitude that you might hope to do that, because the near contribution has a wider distribution in latitude.
Sullivan
So that was your idea? To try to use that.
Schmidt
The idea was therefore, to do a rather conscious effort to try and get the cross-sections perpendicular to the plane with the highest possible accuracy. And the way we did that was by making at a given position in the Milky Way at longitude zero, we would sort of make a tracing, just let the antenna stand there, not move it at all in azimuth or altitude and because of the geometry of things the thing went across at an angle of about 30° from what you wanted. In other words, 60° to the plane, which was a pretty good cross-section. And it was a great detection, of course, to this kind of observing, because you had nothing to do. You could let it go for an hour or two and then you would move it again ahead of the Milky Way and you'd make a cross-section.
Sullivan
And you picked the peak hydrogen velocity, I suppose?
Schmidt
Yes. Well, we picked, I picked velocities all along the entire profile that went with that particular position in the galaxy, so I could as a function of velocity, try to do the disengaging of the two contributions. One hoped essentially, in some cases it was really also observed, ideally one would really like to see a broad profile and then sort of near the middle, where the peak was about to appear, a sharp and narrower peak that was the contribution and the rest was a [?] by one. But, of course, at radio velocities and frequencies where the distance ratio between the two points was not expected to be so large, say a factor of two and half or so, it was rather more difficult to separate it out. But I did succeed nonetheless in doing a reasonable separation and I remember that much of the deduction, which, of course, was again done by hand, much of it by myself and of course with extensive help from people computers. I remember that at the final phase of the whole project that the result essentially came in positions all over the central part of the Galaxy, of course, only the northern hemisphere for different radio velocities or frequencies, distances, of course, the two contributions separated out, nearby and far, and for each of them an intensity or a hydrogen density was finally computed. And I remember that finally I think it must have been in the summer of probably, ‘55, sometime in August, it was very hot on a Tuesday evening or so, I had finally a list of 808 points like that with the position in the galactic density and I could plot it up. And that was really quite exciting because I plotted on a fairly large scale, I plotted all these 808 points and wrote in the density and then started drawing just from nothing, just started to draw iso-density contours, and by gosh...
Sullivan
Had some structure.
Schmidt
You saw structure and you could follow things around half or one quarter of the Galaxy or so at 180° or perhaps 90° or 100°. It was really exciting. Especially to realize you looked at distal-ides in the galaxy that of course, were very large considering what you can do optically. I mean, you looked at things, ten, fifteen kiloparsecs...
Sullivan
It was completely unknown, what was going on.
Schmidt
Yes. And in the interior part had never been seen or disentangled yet. So it was really very interesting.
Sullivan
What was your feeling as to the reality of these features, though?
Schmidt
I thought that they were real; I thought that they were real. And it's interesting, of course, that since that time, that things have not changed an awful lot yet. The interpretation, of course, of these iso-density contours maps has changed somewhat. Like Burton has shown if you have a velocity perturbation rather than density perturbation, by jove, you can almost do a continuum with a uniform distribution. But nonetheless, something is happening that does show a circular or a largish scale spiral structure and that's probably the main thing. That was sort of interesting. Around the same time, I was working on a thesis topic which was to be the mass distribution in the Galaxy, which there was a good incentive at that time, namely, since you can from 21 cm only get the rotation curve out to the solar distance, the main purpose of the exercise was to get the location for outside the [?]. Essentially, by putting together a mass model that fitted well, inside the solar distance, and then to extrapolate that mass model itself by what seemed to be a reasonable distribution of matter and then to compute the rotation curve in the outer part of the Galaxy on the basis of that.
Sullivan
Well, wasn't that inner rotation curve also quite new? I mean...
Schmidt
That was fairly new and was, in fact, I think the first publication of it was only in ‘58. No, that sounds strange. ‘54, it must have been- [K. K.] Kwee, Westerhout and Muller, and that, of course, it was in itself was very exciting to see that happen because all we'd known to that time had been a and b is [?] properties in the Sun's neighborhood and no other information. It was about it did become feasible, you're quite right, it did become feasible at that moment to come up with a mass model that was fairly detailed because the rotation curve for the first time was available fairly detailed rather than just only in overall outline with a local slope that was perhaps fairly good due to an a and b, but we now had much more detail. Until that time, Oort had used units of mass, of matter in his models that were sort of homogeneous ellipsoids of rotation. And at that time, I extended this into the use of inhomogeneous ellipsoids, essentially by taking an infinite number of these ellipsoids of different sizes, superimposed them all together, and it turned out that the formulation, the mathematical formulation on these things turned out to be fairly simple and fairly straightforward. So it was actually not more difficult to work with these inhomogeneous spheroids than it was to work with the old homogeneous ones.
Sullivan
I see. What did...
Schmidt
And that gave nice, smooth results for the models and so on rather than what you would get in this finite number of homogeneous ellipsoids, you would, of course, get sharp breaks in your rotation curve, density distribution and so on, so...
Sullivan
When were you working on this, now?
Schmidt
That was all at the same time, that also started late ‘53, I think and then ‘54, and then I think it was all completed in early ‘56, I think, in March, ‘56 when I got my degree, or May or something. So those two things sort of went on in parallel fashion.
Sullivan
So if it hadn't been so mathematically tractable, then it probably would have been impossible at that stage to...
Schmidt
Well, one would have worked with these homogeneous things which...
Sullivan
You would have fallen back on that.
Schmidt
Which is somewhat messy, it's not so elegant, but it worked out very well. So there was sort of an interchange, I mean, for the working out of this spiral structure in the interior parts, I used these results that had already been obtained for the n point of the profiles in the interior part of rotation curves, Kwee, Westerhout, and Muller. You know, the interior as it were, I constructed galactic models which Westerhout needed for his outer spiral structure studies to convert angle of velocities that he, of course, got from observations into distances to the Sun. So we worked quite closely together all the time.
Sullivan
So you got your degree then in ‘56, you say.
Schmidt
I got my degree in ‘56 and then I got a Carnegie Fellowship at Mount Wilson in Pasadena and during the two years that I was there, almost three years, I did some observational work on galactic clusters, some of which didn't work out too well and I also became interested in star formation and in the consequences of star formation, gas being used up and therefore as a function of time, there being less and, less gas in the galaxy. The gas mass decreasing, the star mass increasing, the...
Sullivan
Was this, do you think, motivated by the working with all this gas and so forth?
Schmidt
It was motivated, essentially, by the 21 cm stuff, although it really got quite a boost from Sidney [Vandenberg?] whom I met in the middle of ‘57 at an AAS meeting at Harvard, who asked me about the Dutch results on all this work on the Galaxy and who became then suddenly very interested in effect that the total gas mass in our neighbor is only of the order of 10%. It means that much of the activity in terms of star formation or in the conversion of gas into stars in effect is already behind us. So he started to work on that, and put up a sort of initial formulation and when I saw that, I liked part of it and I did not like another part of it, so I started to work on some of it myself. I got rather involved in that for the next few years.
Sullivan
Did you know at all about the Harvard, it was just going then with observations of solar clusters. [T.] Kochu Menon did some and...
Schmidt
Yes, yes. That was around that time that was happening.
Sullivan
Did that influence you at all?
Schmidt
No, no. No, no. It was just the overall distribution of gas throughout the Galaxy, which by the way, didn't only show that of course, over here that the percentage was fairly low in mass of gas, only 10%. But also if you go to the interior part of the Galaxy here, in our Galaxy that well, that the gas density remains more or less constant while the model, of course, showed very clearly that the total mass density increases quite steeply. So the sectional contribution by the gas becomes less and less when you go to the interior part of the Galaxy.
Sullivan
Right. And that was first recognized in our own Galaxy before any external galaxies were done.
Schmidt
That's correct. That follows essentially from the work on spiral structure and so on. And that gave rise in 1957 or ‘58, ‘59 to the idea that I expressed at that time that star formation might go as a power of the gas density, which is more than one. In which case, things sort of go that way. But of course, these days, as we heard in the meeting this morning, one can now probably think more physical because it was not a physical law. It was just a feeling that...
Sullivan
You just wanted some strong...
Schmidt
That's right. As we know here, it may have to do with the crossing of the spiral waves through the neighborhood, which is, of course, much more effective. Of course, how do you get it to condense otherwise anyhow.
Sullivan
Were you aware when you were at Palomar of the radio work going on there, the very beginnings of it?
Schmidt
Yes.
Sullivan
In fact, the first HI observations were done on the Palomar Mountain itself.
Schmidt
That's right.
Sullivan
Were you involved in that at all?
Schmidt
No. There was a 30 foot antenna that was on Palomar, the first, while Bolton was here. But of course, the Owens Valley Radio Observatory was had already been started and it was, I think, around ‘55 or so, '56. No, it’s rather later. I shouldn't talk about the early days of Owen's Valley because...
Sullivan
Oh, I have talked to Gordon Stanley.
Schmidt
Yeah. But, of course, he interesting one that was going on at that time was done by [Thomas "Tom" A.] Matthews and [P.] Maltby and [Alan T.] Moffet when they found the double structure of many of the radio sources, and that followed the work of Jodrell Bank where [Roger C.] Jennison had shown that Cygnus A was double. Now I did not do anything really active in radio astronomy during that first visit here in 1956-58. I worked mostly with the optical astronomers; had contact with [John G.] Bolton and Matthews and so on. But I wasn't active in radio astronomy. Then when I came back in late ‘59 to take up a position at Caltech, it took about a year or so, year or year and a half, at a time when [Rudolph] Minkowski had retired. Minkowski retired in June, 1960. And just in his last run in May, 1960, he had been able to get 3C 295 with a redshift of 46%, which until really recently still was the record- the remarkable achievement of his. And then when Minkowski retired, it was clear that the radio astronomers were doing such interesting work and identifying these elliptical galaxies that work ought to go on in terms of the redshifts. It was very badly needed. While at one time, some of us set together and felt that each of us could once in a while take one of these galaxies that was probably not so technical and I got somewhat more interested and together with Matthews who at that time, it was 1961 about, the position and identification work. I started to fairly regularly from 1961 onward to observe radio galaxies to get directions. And it was sort of pleasant work.
Sullivan
I wanted to ask you, you felt then that you were sort of carrying on in Minkowski's, if no one did it, then it would really lag...
Schmidt
Yes. Because it was clear that it was very interesting and the redshifts were very badly needed. Oh, when I took over from Minkowski, I think that a redshift of only 10 or 15 radio galaxies were known, and almost all of those were very bright, you know, like Centaurus A and Cygnus A and M87 and so on. And I started on the very much fainter galaxies then, which were the sorts of observational challenge, which turned out to be feasible, and found that most of those faint ones actually did show emission lines that were quite strong and that was interesting work and it was do-able and I made it my program.
Sullivan
Let me ask you about the relationship between the optical astronomers and radio astronomy. It's been my impression that there were very few in the ‘50s who properly appreciated what radio astronomy was developing into.
Schmidt
Yes.
Sullivan
Do you have any opinions as to why that was?
Schmidt
No. Well, perhaps yes. It was mostly, I think, a lack of communication. For one thing the people in the new field and the people in the old field, always, of course, talked a different language, and at that time, you might also say that was the first time that a new, I think that's right, the first time in a long time, that a new wavelength region was opened up. And it was opened up by people who were mostly engineers, radar engineers from the war. They talked a rather different language. So the astronomers, in many cases I think, simply didn't quite understand what they were talking about. The astronomers themselves talked their own language which is not so simple, it has lots of gobbledygook and expressions and things that do not help, certainly, to communicate things to engineering type people. And I think it led, initially, to not much communication, and it was only a few people, I think, in optical astronomy who saw across these differences and realized the basic interest of that. Jesse Greenstein, for instance, was immediately interested in radio astronomy.
Sullivan
Oort...
Schmidt
Oort and Minkowski, too, and [Walter] Baade. But many astronomers simply did not appreciate what was going on. I think that is very different these days, I mean...
Sullivan
Well, perhaps after radio had done this, then they were willing to accept X-ray astronomy when they came along.
Schmidt
That's right. That's right. And these days, isn't it right, if somebody talks a different language, you cross the barrier and try to find out what he was saying and you yourself talk a rather more, here, about astronomy than you used to. You saw this astronomical terminology that is rather dense. So I think things have improved a lot. But in the beginning, yes, there was little communication, little understanding, I think. But because of Oort in Holland, they didn’t suffer from it.
Sullivan
That is the one exception.
Schmidt
Right. So this work went on and I think it was in 1962 that...
End of Tape 37B
Sullivan Tape 38A
Sullivan
This is continuing with Maarten Schmidt on 18 August ’75. So in ‘62, you say, you became interested in star-likes?
Schmidt
It was the middle of 1962 that I worked on my first star-likes object, which Tom Matthews had identified and given to me, but of course, before that, late in 1960, 3C48 had been pointed out as a star-like identification of the radio source by Tom Matthews. In fact at that time, they would have been called star, it was thought to be a star in the Galaxy.
Sullivan
Oh, I see. Was that term coming back into vogue at that time?
Schmidt
Oh, yes.
Sullivan
It'd sort of gone out for a while.
Schmidt
Yes, it had gone out, because...
Sullivan
It was inside the galaxy.
Schmidt
There were no stars at that time that were radio sources, it had to come out but then with 3C48, it was locally reintroduced again, as you know 3C48 was found by Matthews and [Allan R.] Sandage and Bolton and [Munch?] and Greenstein all together. I happened to be not involved in it at all, but late in 1960, to be a variable star with a blue color, some emission lines, and it was right on the position of the radio source of 3C48.
Sullivan
What were the error bars in the radio position at that time roughly?
Schmidt
I'm not sure. I think it was, at that time, a rather accurate one, and it's a rather bright source, so let me guess something like better than 10 seconds of arc.
Sullivan
I can check that. It would have to be very good, I guess, before you'd have any confidence.
Schmidt
Between five and ten seconds of arc, I would now guess. But I don't know. It can be checked, of course. Somehow that thing had gone by me. Tom had simply been involved with others on that particular thing, and I hadn't given it much attention even when it happened, as can sometimes be the case. I was involved in my galaxies. But when he had another, 3C286, I observed it in early ‘62, I think, or so. And I found that it had one emission line and I simply couldn't understand what it could be, 5170 angstroms where nothing shows anything, no object ever shows just the line at 5160 angstroms and in fact, I wrote a very short letter to the Astrophysical Journal only this long, about three inches long, a sort of complaining note, it's the first one about quasar detecting, that was 1962, in which I say that I have this object spectrum and it doesn't look like anything we have seen before. Because 3C48 had not been published yet, it had been at the AAS Meeting, but the publication was very badly held up until ‘63.
Sullivan
So you're still thinking in terms of a star?
Schmidt
Oh yes, at this time, I was just thinking of a star. Well, that went on and Tom soon came up with another case, which I didn't publish, and which showed no lines at first, 3C196, and Tom came up with another star again, all in the course of the next, one and a half or the next year or so. 3C147 which showed two lines; these were different positions than, different wavelengths, than the one line was 3C286. And then of course, came...
Sullivan
Let me just ask you, did none of these show any wisps or anything? Does 3C48 have a wisp is my memory correct?
Schmidt
Yes, 3C48 has a wisp.
Sullivan
Did you notice that at that time?
Schmidt
They had noticed it. It was theirs, that was the one done in 1960, you see, 3C48, yes, they had noticed that.
Sullivan
They didn't make anything particular of it?
Schmidt
No, no, because it's quite difficult to observe. It has only been observed properly quite recently.
Sullivan
But didn't make anything of it as far as what it might mean that it's not really a star?
Schmidt
No, no, no. At least I'm not aware of they felt that way. But anyway, when I say that these other stellar objects came through in 1962, I seem to be getting into some time trouble. It must all have happened mostly in 1962, yes. Because the first spectrum I took of 3C273 for which the position came from Bolton, Matthews and from the occultation observed in Australia, by the moon, by [Cyril] Hazard and others. I think I observed that that thing for the first time very late in 1962, around December 28th or so, and this thing was so bright, it was 12th or 13th magnitude. It was so bright it completely over exposed the first , even though [?], it completely over exposed it. But I do remember that right at the end of the spectrum, near the atmospheric limit, near 3200, the spectrum seemed to cut off very sharply and when I looked more, it was totally over exposed you see, but it cut off very sharply there. When I looked more carefully, there seemed to be a line there. So anyhow, I took other, better exposed spectrums of it soon, for the next few nights, and these showed up something like five or so emission lines that all that funny places. None of the at them at positions in wavelength of the earlier star-like objects, so it remained totally unclear.
Sullivan
What was the most lines that you had before for these objects?
Schmidt
Well, two in 3C147, two or four in 3C147, but are rather close together, I think only two of them finally turned out to be real. One in 3C286; I don't know what they had in 3C48 - they already earlier had quite a few lines, like five or six or so.
Sullivan
So it wasn't unique in that respect?
Schmidt
No, no. Now in hindsight, I really treated this thing rather naively, I must say, in the sense that I told everybody. I mean, I came back to Pasadena, I measured the wavelength and I told everybody that I had this whole slew of lines that I couldn't identify, and I remember Dr. [Ira S.] Bowen, who at that time was Director of the Observatories, had a hand in it and came up with some very tentative identification with some helium lines. But it didn't look good, nonetheless. Dr. [Baschek?] from Kiel, who was there at the time visiting, made an attempt and you know, in hindsight, it was a bit silly almost to do this thing this way. In telling everybody that you have these five lines and what did they think they are, you know, if you knew what it had been, it would seem kind of silly. But anyhow, in the next observing room that he had, [?] worked with his scanner in the farther [?] that I couldn't easily reach and found a line at 75, yes, at 7600, one single line, a strong one. And well, that was added then to the list of lines that people heard about and there was no really, not any real progress then. And then came a letter, and that must have been late January or so of 1963, about a month later, only, because I had immediately written Bolton in Australia, who sort of was my contact man with Hazard whom I didn't know at that time. He said, "Cyril Hazard is going to write the thing up for Nature, why don't you perhaps write a note that can be published next to it? So that we can have the whole story in one gulp."
So I thought that might be a good idea, and I think it was on the 5th of February that I sat down and in the afternoon after lunch on a Tuesday, and I sat down in my office and I was to write that letter. It could be pretty brief since I didn't understand things. And I remember that I looked up something on the spectra, I wanted to check something, so I had a bunch next to me and I looked at them, and then rather suddenly it struck me that there was a certain regularity to the lines. If I left out two of the lines, there were at that moment six lines, I think, and if I took the one infrared line, plus three of the visible ones and forgot about two of the visible ones, it wasn't all that trivial, that those four were left seemed to be regularly spaced decreasing space in going toward the blue and getting fainter and fainter. Now for some stupid reason, I decided to make an energy level diagram, which in hindsight, is totally unwarranted under the circumstances, you see, so I computed something and I must have made an error or so because things that I then wanted to work out didn't fit at all. And I got essentially annoyed, and said, "But there is regularity, I see it." You see, the error I'd made seemed to say that there was no regularity.
Sullivan
But why do you say it was not warranted to make an energy level diagram? That seems like it would be the thing many people would do.
Schmidt
Well, anyhow, I did it. But what happened then, is that I decided in order to check the things were really regular I decided to take the ratio of these lines to the neighboring Balmer lines. So I took the line in infrared, I took the ratio to H-alpha is 1.158 and I took the line at 5600 and took its ratio to H-beta and it was 1.159 or so. And I looked up and took the next ratio, it was 1.158 again, and the last one was also something like that.
Sullivan
And the only idea of this was to check the regularity which you thought you saw with your eyes?
Schmidt
That's right.
Sullivan
And that gave you the redshift.
Schmidt
That was a shock, of course. So that gave me the red shift. And then in the next few minutes, I checked what the rest wavelength of these other two lines that I'd thrown out would be, and the one fitted beautifully with O3, 5,007 and the one in the ultraviolet at 3200 fitted very well with Mg II at 2800 angstroms which had never been seen yet, but which I think I knew that it might be expected. So that seemed a rather positive confirmation of the idea that this was a correct identification. So I sat there for a while stunned. Because a 13th magnitude star doesn't have a 16% red shift, you see, that doesn't happen. And now I perhaps paced up and down the corridor and I saw Jesse and I said to him, "Come and listen and look, I mean, hold on to your seat, you won't believe this." So he mulled it over for a while and then Jesse said after ten minutes or so, "Well, why don't we look at the lines of 3C48 again? With that in mind." And it took not more than two minutes to then get the red shift of 3C48 at 37% or 36% or whatever it was.
Sullivan
Why did you think you had not seen it in 3C48 first because that data had been around for a while, hadn't it?
Schmidt
Well, I had never looked at that, of course.
Sullivan
Oh, that wasn't your data, that's right.
Schmidt
No, that was the other one; that was the first one found. I never looked at the data. And also, it wasn't so obvious in it because the red shift by that time was so large that things had really shifted a lot. By about 1000 to 1200 angstroms, which is quite a lot. Well, we again found the 2800 line then, something that appeared as 3800 we then identified with a 36% red shift as being the 2800 doublet or line from Mg II again, which sort of was a mutual reconfirmation of this thing. So things fitted together very well and we spent the rest of the afternoon essentially checking whether there was any chance that you could get, that this could perhaps not be a red shift but could be some kind of funny ion or so. I mean, we worked that out and by the next day that was worked out, it couldn't be the case. But by that time, things became rather convincing.
Sullivan
Were you not worried about the calibration problem?
Schmidt
No. I mean these wavelengths, of course, were very well established, isn't that right?
Sullivan
No...
Schmidt
[?]
Sullivan
No, I mean that your standards had gotten off somehow.
Schmidt
In the spectrogram?
Sullivan
Yes.
Schmidt
No, no. No, no. Well, that was that. To cut the story somewhat short and in order to wind up at 1965, what really happened after that time as far as I was concerned, was in effect, to my mind, up to 1965 was a more interesting and as far as I'm concerned a more satisfying, further development than the discovery of the red shift itself was. And that was the extension of the red shift to very large ones of the order of two. See, it was clear that if we can have for a 13th magnitude object, a 16% red shift, then for an 18th magnitude object, you should have a red shift of the order of 10 times as much, therefore, very large red shifts were foreseen on that basis.
Sullivan
It was the Cygnus A thing all over. As soon as you knew you had Cygnus A, you knew that the others...
Schmidt
Yes, that's true, but here the shifts in wavelength were so enormous that you now had to base your line identification which are always done between, say, 4,000 and 6,000 angstroms in the normal photographic region, we now had to, in order to get the very large red shifts, like two, you had to go to a rest wavelength between 1,000 and 2,000 angstroms where no solid information about what lines should come up was available at all. And it became sort of a process in which objects were found with increasing red shifts such that once in a while an object would show a line in the far ultraviolet, near 32-3500 angstroms.
Sullivan
And then you could use that...
Schmidt
Then one could, it was sort of a good step as it were, in which you pull yourself over farther and farther to the ultraviolet by once in a while identifying a line which appears in the ultraviolet for a higher red shift source, and then that line can be used to get even higher red shift. It's easier said than done, though, to just ask for a higher red shift source, because out of all the star-like objects that Tom Matthews started to produce, one had no idea which ones were the ones to, in which order you should work on them. And also, these things were very faint in general, on the order of 18th or perhaps somewhat fainter in magnitude at that time, the typical exposure for the spectrum that could be used to get these lines in red shifts was in the order of 5 or 6 hours, so on a typical run through Palomar I might sometimes, including perhaps not too good luck with the weather once in a while, might often come back with three or four spectra. And then it was not necessarily an object you were so interested in because it might have quite a long red shift. It might even be so large that you couldn't, you know, decipher it yet or it might be so small that it didn't produce any new lines.
Sullivan
Did you make anything of the wisp in 3C48 once you had established that as having a...
Schmidt
Well, no. It was argued by Matthews and Sandage when they published their paper, finally in 1963, taking into account that the red shift was 30 some percent, it was argued by them at that time that the thing was too bright to be a galaxy, but now very recently, of course, it's been shown to be an HII region and so it is not a normal sort of galactic-like thing. And so, that's sort of in a nutshell, I guess, what happened to me from ‘54 to ‘65. I've really been most active, I think, in the optical astronomy things, but almost all the time it's been closely tied in with radio astronomy. Right now, I'm doing quasar surveys and get all my basic information from identified radio sources really, star-like ones.
Sullivan
Let me ask another question. When you had the ratios, it seemed like you immediately said, "Well, that's a red shift," but it might have been something else. It could have been a scattering.
Schmidt
Yes. We discussed it at a very early point. In the next day or so we started talking about it locally, our own physicists at Caltech urged us that we should particularly investigate the gravitational red shifts because that, to them, seemed like the most likely thing. But all these things had forbidden lines, you see. 3C48 had strong forbidden lines, which means that the density of the gas emitting the lines had to be very low like 104 and when you have a gravitational red shift, the linear size of the total thing must be very small unless you're an incredibly large mass. M/i you want to be very large, and therefore, r has to be very small. Therefore, the emitting region is very small, but if the gas density in addition is small, you have a lousy radiator in the lines, and there is no way to put it at a decent distance. So...
Sullivan
So you eliminated that.
Schmidt
So Jesse and I discussed it at a quite early stage and managed to, well, in the first instance dismiss it and at a later time, I re-discussed it and it now looks even worse essentially. Of course, the thing that's happened in the meantime is that some people have argued that the red shifts are not so much gravitational, but rather that they are simply non-cosmological. I mean, a number of people take the point of view that there is something at work that we don't know about yet.
Sullivan
That came in somewhat later, though.
Schmidt
That came in somewhat later, that started around ‘66 a little bit, but much more around ‘69 or ‘70 by the work of [H. C.] Arp and [Geoffrey R.] Burbidge.
Sullivan
Okay, well, thank you very much.
Schmidt
You're very welcome.
Sullivan
That ends the interview with Maarten Schmidt on 18 August ’75 at San Diego.