W. Heisenberg: How certain is it that the spirals are perma-
nent structures? May it not rather be a process of continuous
formation? Spiral structure might be very quickly washed out
by rotation, but new spirals could be formed by fluctuation
J.H. Oort: I agree. But it is difficult to conceive how spiral
structures which extend over an entire galaxy could be formed
entirely anew at intervals of one or two revolutions of the
Oort 1965, p.23
...then a spiral arm is some sort of a wave. Once one says
this, of course, one runs into an enormous number of possibilities.
Prendergast 1967, p.304
1.1. Cambridge union
Since Lord Rosse discovered spiral structure in M51 the expla-
nation of this beautiful form has been one of the outstanding
problems of cosmogony. The straightforward belief that this
structure is a natural consequence of a swirling motion was
probably held by many of the early observers and it is our
hope that the present work goes some distance to establish
that belief on a firm theoretical foundation.
Goldreich & Lynden-Bell 1965b, p.125
Donald Lynden-Bell and Peter Goldreich met in the fall of 1963 in Cambridge, UK. One of them had been back there from the USA for a year already, as a University lecturer in mathematics and director of mathematical studies in Clare College, 3 the other had just arrived on a one-year US National Research Council postdoctoral fellowship. 4 Goldreich had a concept of actual problems in galaxy dynamics and no plan to pursue them, but Lynden-Bell got him captivated by the prospects of spiral regeneration. 5 Enthusiasm and youth - Lynden-Bell was 28, Goldreich was 24 - tempted the Cambridge researchers by the confidence that this trail would lead to the solution of the old great puzzle, and they started marching on the "spiral arms as sheared gravitational instabilities" (Goldreich & Lynden-Bell 1965b, hereinafter GLB) with a salvo of "requirements of any theory".
"Any theory must be wide enough to contain the bewildering variety of galactic forms. The conventional picture of two spiral arms starting symmetrically from the nucleus and winding several times around like continuous threads is wrong in several aspects. In only about a third of all normal spirals can it be claimed that just two arms are dominant and although in these there is some tendency to symmetry it is not always very pronounced. [...] The remaining two-thirds of normal galaxies are multiple armed structures. In Sc's the arms often branch at unlikely angles and the whole structure is considerably more messy than the conventional picture. A swirling hotch-potch of pieces of spiral arms is a reasonably apt description. A correct theory must have room for neat symmetrical two-armed spirals, but it must not predict that most normal galaxies should be like that. The mechanism of spiral arm formation must be so universal that it can still work under the difficult messy conditions of a typical spiral galaxy" (GLB, p.126).
No less categorical was the authors' view of the acute `winding problem' raised just a few years ago (Prendergast & Burbidge 1960; Oort 1962) to strengthen the evidence that "anything in the Galaxy is sheared at such a rate that at the end of perhaps one or two rotation periods it will be quite unrecognizable" (Prendergast 1967, p.304).
"Unless the galaxies have conspired all to be spiral together for a very brief period we must deduce that either (1) the spiral structure rotates nearly uniformly although the material rotates differentially, or (2) the arms are short-lived but reform as open structures, or (3) that the observations are wrong and spirals rotate nearly uniformly" (GLB, p.127).
"To admit (3), is to say that the theorist is bankrupt of ideas", - GLB judged (p.127); definitely higher they favored "perhaps the most promising of the theories based on (1)" that was being made across the Atlantic (Lin and Shu 1964), 6 yet what they found even more consistent with the sheer complexity of actual galaxies was their own "second type of theory".
Spiral arms, the authors reasoned, are recognized above all by their brightness due to hot massive stars that are being formed there. For all that formative period, considerable compression of interstellar gas is needed. It logically calls for Jeans instability as occurring mostly in the spiral arms, and "this at once raises the question whether the arms themselves can be due to gravitational instability on a slightly grander scale" (GLB, p.126).
3 Once he got his PhD degree in 1960, Lynden-Bell left Cambridge for California. Working at Caltech with Sandage, he solved problems on isolating integrals of motion (see Lynden-Bell 1962) and also made, with Sandage and Eggen, the classical work on high-velocity (old) stars (Eggen et al 1962) that proved the fact old-time contraction of our Galaxy. Besides, he visited Chandrasekhar at Yerkes Observatory for large-scale instability problems. Back.
4 Goldreich's 1963 thesis was on planetary dynamics. Back.
5 "What I knew about spiral structure I learned from a course at Cornell entitled "Cosmology and Evolution" that I took in the winter of 1962 while still a graduate student. It was the only astronomy course I ever took. (After completing my thesis, I was appointed an instructor and taught the course the following year.) From this course I learned that young stars were concentrated in spiral arms and became aware of the winding problem. [...] My thesis advisor Thomas Gold mentioned Donald Lynden-Bell's name to me as someone who did interesting work on stellar dynamics. Otherwise I didn't know anything about him before arriving in Cambridge. Nor did I have any intention of working with him. I cannot recall how and why we started to collaborate, but probably it was due in large part to Donald's infectious enthusiasm for pretty much any topic in astronomy or related fields". (Goldreich)
"I think my enthusiasm was that the stability of a differentially rotating disk even one modeled as gas had not been worked out and understood and our mathematics should allow us to understand that problem". (Lynden-Bell) Back.
6 It is not entirely clear when and how GLB had first learned about Lin's spiral interests and initial steps. Lynden-Bell does not think they had "any thoughts about Lin or about steady waves" when they worked in 1963 (Lynden-Bell). But soon afterwards they knew about the Lin & Shu 1964 paper from its preprint that Lin had sent to Lynden-Bell in mid-July 1964 to acknowledge his own receipt of the GLB preprints. "My reaction to that paper was that Lin and Shu had missed out the real problem by leaving out the pressure. While I read that paper my feeling was that had I been sent it to referee I would have rejected it. [...] I believe that if the paper of Lin and Shu had not been written we would have written essentially the same paper, and I think [one has] the information to deduce that from [...] my thesis along with our GLB paper, and one is a natural outcome from the other and the more detailed stability calculation." (Lynden-Bell) Back.