by Jaan Einasto
The personal computing industry developed very rapidly, and in the middle of the 1980’s we had to think which computer to buy next. The Apple Lisa was expensive, moreover too few software were available for Apple computers, so we decided to switch to the IBM PC line. During the trip to Nordita in 1987 I was able to spare enough money to buy a PC. The development was so rapid that in my next few visits to the Institute of Astronomy I bought a new PC each time. In 19901 was visiting ESO, and in the computer shop Vobis I saw a not too expensive notebook. The development of notebooks was also very rapid, thus from this year on in almost every year I sold the old one and bought a new, better notebook model.
In the early 1990’s our team got a grant from the Soros Foundation. We used the grant to buy personal computers, so soon all our team members had their own PCs. Also it was possible to buy for the Observatory better computers; in the mid 1980’s we got our first UNIX computer. Since then we used for scientific computations first UNIX, then later Linux operating systems. Linux was available also for personal computers, thus most of us switched to Linux. However, there are a lot of programs for non-scientific use, and most of these programs were available only for the Windows operating system. So most of our Linux users had two operatings systems installed, Linux for science, and some version of Windows for everyday life.
About the mid 2000’s we discovered again Apple computers. The operating system of all Mac computers is UNIX based, thus all the advantages of Linux are available. But Mac has also programs for almost all everyday applications, many of them the same as for Windows. The user interface of the Mac operating system is much more user friendly than that for Windows. So many of our team members are now using Mac computers. I too have used for about 6 years only Macs — MacBook Pro and MacBook Air for computations. My iPhone is not only for communications, but it contains almost the whole of my music collection, and is a good camera always with me.
4.4.2 Life in the Observatory
In the early seventies the Estonian Academy of Sciences planned to divide the Institute of Physics and Astronomy into two institutes. According to the plan the present Institute was renamed as the Institute of Astrophysics and Atmospheric Physics, and a new Institute of Physics was formed. But here there was a problem. According to rules accepted in the Soviet Academy of Sciences, the number of Doctors of Sciences was of particular importance. In the Soviet system a Doctor of Sciences degree is approximately similar to Doctor Habilitatus in Germany. To get a Doctor of Sciences degree usually about 20 years of hard work was needed, as well as a large number of publications. The number of doctors determined the rank of the institute and thus the size of salaries, and the financing of the institute in general. Since Aksel Kipper was the only Doctor of Sciences among our astronomers, doctoral defences became vital.
We all expected that Grigori Kuzmin should defend the Doctor of Sciences thesis, but he was too reluctant to do this. How to encourage Kuzmin to compile his doctoral thesis? Kuzmin usually achieved his results through a rather brief, yet very intense period of thinking and analysis, but compiling the results into an article took him a lot of time. Several of his results stayed in the drawer because he could not pull himself together to write the article. It was likewise with his doctoral thesis — he could not mobilise himself to write it, arguing that he was more interested in mulling over obscure problems rather than drawing up long volumes that contain nothing new. So what about the thesis?
I discussed the issue with my younger colleagues and we found that his published works were so good that they could be used as chapters of a solid thesis. Having decided this, we announced to Kuzmin that we were going to start putting his papers together to form a Doctor of Sciences thesis. Due to the requirements of that time, this meant re-typing of all papers. After some resisting, Kuzmin agreed. But as he reread the old articles, he come up with new ideas to be used as supplements to the chapters. How this transpired: Kuzmin was up well into the night, writing his additions, which were on the table by the morning for us to somehow make out the garbled handwriting and type them in. Around noon the Maestroshowed up again and examined the text, adding copious new improvements that we typewrote once more. The process was quite effective and after half a year, a unique piece of work was completed. Most of the additions have unfortunately still not been published, but the defence of the thesis was a great success, and once again, Kuzmin’s opponents had to acknowledge the exceptional quality of his work.
Almost all Kuzmin’s papers were published in Russian, mostly in Tartu Observatory Publications, some in Soviet journals. So his results were not known to the English speaking astronomical community. About ten years ago one of our young collaborators, Peeter Tenjes, translated Kuzmin’s most important papers into English, including his additions to the doctoral thesis. Recently we had a small conference to celebrate the 200th anniversary of the Old Tartu University Observatory. One of the main speakers was Tim de Zeeuw, who gave a review of Kuzmin’s work. We discussed the publication of English versions of Kuzmin’s papers, either as a special issue of thejournal “BalticAstronomy”, or just on arXiv.
After the successful defence of the Kuzmin’s doctoral thesis, I decided to use the same method with my own thesis. Most of the chapters comprised of earlier published works. I added several new chapters using my latest results on the evolution of galaxies and the new model of our Galaxy. The defence took place in March 1972, and went equally well. Regrettably, my newer chapters also remained collecting dust. New projects kept me busy, and several pioneering works on the evolution of galaxies are so far not published. Most of my later publications are already written in English, so I am now preparing English versions of these chapters to put onto our website.
By the end of the sixties our team was rather small. However, the studies of the structure of galaxies, including difficulties with the explanation of the dark matter problem, had reached a stage where a need arose to hire young astronomers who could examine the problems more thoroughly. According to our development plans, all the young astronomers would in the 1960’s go to work in the field of stellar astrophysics. Now it was time to start the second phase of the plan and to devote more attention to galactic studies. Grigori Kuzmin and I turned to Kipper with the request for permission to start preparing students for studying galaxies. To our surprise, Kipper stated that if we were to present the plan to the board of the Institute, he would oppose it. We were quite astonished — the plan of the development of the observatory was approved by the director himself. But we did not wish to cause frictions, so we decided to work towards achieving the plan, but by going about it discreetly.
We made presentations of the new problems concerning the structure of galaxies to physics students and encouraged them to choose the structure of galaxies as their subject field. Soon we had new young astronomers among us after a long time: Erik Tago, Jaan Vennik, Ants Kaasik, Peeter Tenjes and Peeter Traat. Another problem was in our education. The older generation of astronomers in our group, including me, had a classical astronomical education. But to solve new problems a good knowledge of theoretical physics and cosmology was needed. Thus we started collaboration with Enn Saar and Jaak Jaaniste, who had excellent education in theoretical physics and were already working in the department of theoretical astrophysics.
The study of galaxies soon became our main course, taking precedence over classical problems of stellar statistics and dynamics, performed by Grigori Kuzmin, Heino Eelsalu, and Ülo Veltmann. Mihkel Jõeveer switched to more practical problems of stellar dynamics, in particular to the study of dark matter and the structure and distribution of galaxies.
It was customary to write for the Academy reviews of the results of the past five-years, and to highlight the most significant results. In the mid 1970’s again a five-year period was over, and highlights were discussed in the council of the Institute. There was a dispute on whether solving the problem of dark matter could be mentioned among the highlighted results. Kipper opposed this suggestion; according to his
opinion our results were too “populistic”.
One more episode from the 1970’s. At the IAU General Assembly in 1970 in Brighton I met Gerard de Vaucouleurs. We discussed problems of the structure of galaxies — he is one of the best experts in this field. He was interested in our models of galaxies and offered cooperation. Several times in the 1970’s he sent me invitations for a visit. At the time it was very difficult to get from Soviet officials permission for such visits, in particular to Western countries. Before the Soviet Academy of Sciences applied for a visa, approval from a number of intermediate instances was needed. The very first instance was our own Institute, thereafter various organs in Tartu, Tallinn and Moscow. Once, when I started the application, director Kipper invited me to his office and suggested I withdraw my application since this could block foreign visits for our stellar astrophysicists. So, the cooperation with Gerard was not realised. Only many years later, when I received an invitation from George Abell to discuss problems of the distribution of rich clusters of galaxies, I had a chance to make a short visit to Austin and to meet Gerard.
These episodes do not diminish Aksel Kipper’s role as the founder of the new observatory. He was clearly one of the best scientific organisers in Estonia in the Soviet era. The above-described events show Aksel Kipper’s human side — even the best of us have our shortcomings. The way Kipper viewed science was shaped in the thirties, when the most modern branch of astronomy was the physics of stars — their energy sources and evolution. The study of galaxies was not considered as a branch of astrophysics at the time; observational cosmology in the modern meaning had not formed yet. Despite his somewhat skeptical attitude towards the newer directions in cosmology, Kipper never hindered their development. This is how he once expressed his mentality towards the existence of different directions in the observatory: “Let a hundred flowers bloom, for we cannot foresee, which blossom will come to bear the best fruit.” This is a stance one does not meet everywhere.
Chapter 5
The cosmic web
In this Chapter I shall describe how our team changed our main research goal from galaxies to systems of galaxies. This was a natural extension to our earlier work directed to better understanding of the structure of individual galaxies. This happened in the early 1970’s and continued in the 1980’s. We had so far little experience in the study of the distribution of galaxies because almost all our attention was directed to galaxies, their populations and the evolution of galaxies. I start the story with a description of the circumstances that brought us to the study of the distribution of galaxies.
5.1 Early studies of spatial distribution of galaxies
As described above, according to the classical cosmological world view, based on the study of the 2-dimensional distribution of galaxies in the sky, most galaxies belong to the general field, and only a relatively small fraction of galaxies is located in clusters. These studies suggested that field galaxies are distributed more-or-less randomly.
One of the first hints of difficulties in the classical paradigm came from the study of the distribution of a homogeneous sample of Sc I galaxies by Rubin et al. (1973, 1976a,b). Rubin compiled an all-sky sample of Sc I and Sc II galaxies in apparent magnitude interval 14.0 ≤ m ≤ 15.0. The authors found that the distribution of redshifts of these galaxies is curious. In one large area of the sky redshifts of galaxies cluster around a value 6,400 km s−1 but in another large area redshifts are clustered around a value 4,950 km s−1. Areas of different mean values of redshifts are approximately located in opposite regions of sky, thus Rubin et al. suggested that one possible reason for this anisotropy may be a large motion of the Galaxy and the Local Group with respect to the general field of galaxies.
I discussed with Mihkel Jõeveer these results and we tried to understand the reason for the anisotropy of redshifts. At this time we had already started to investigate the large-scale distribution of galaxies and clusters of galaxies (see the next Section), and had a catalogue of Zwicky near clusters. Our study indicated that Zwicky near clusters are located very inhomogeneously and form large superclusters. The area of the sky where Rubin found lower redshifts of galaxies contains one of the largest nearby superclusters, the Perseus–Pisces supercluster. The area with larger redshifts contains the Coma and Hercules superclusters (Einasto et al., 1975a). These differences in mean redshifts of galaxies could be explained if galaxies also cluster into superclusters, similarly to Zwicky clusters. The number of galaxies with known redshifts was in the mid 1970’s still rather small, thus our result was tentative. But it suggested that superclusters are not just clusters of clusters; here both galaxies and clusters of galaxies of various richness form density enhancements.
Chincarini & Rood (1972, 1975, 1976) measured redshifts of galaxies in an area close to the Coma cluster and found that redshifts of galaxies are concentrated around three distinct values: about 1,000 km s−1,4,000 kms−1, and 7,000 km s−1. Galaxies of the first concentration evidently belong to the Virgo supercluster, and of the third concentration to the Coma supercluster, while the intermediate concentration consists of the N4169 group of galaxies and some other nearby groups. The space between these concentrations does not contain any galaxies in the magnitude interval used for the study, mp ≤ 15.1. The Coma supercluster is detected at a radial distance from the center of the Coma cluster of 14.2 degrees; there is no evidence for the existence of a homogeneous field of galaxies between these three concentrations (Chincarini & Rood, 1976).
The study by Chincarini & Rood (1976) is one of the first indications for the presence of voids in the galaxy distribution. A more detailed study of the environment of the Coma superclusters by Gregory & Thompson (1978) has confirmed results by Chincarini and Rood.
5.2 The discovery of the cosmic web
5.2.1 Zeldovich question
After my talk at the Caucasus Winter School in 1974 on dark coronae of galaxies Zeldovich turned to me and offered collaboration in the study of the Universe. He was developing a theory of the formation of galaxies (the pancake theory); an alternative whirl theory was suggested by Ozernoi, and a third theory of hierarchical clustering by Peebles. Zeldovich asked for our help in solving the question: Can we find some observational evidence which can be used to discriminate between these theories?
Fig. 5.1 Yakov Zeldovich with his wife visiting Estonia, late 1970’s (author’s photo).
Later I heard from my Moscow colleagues that Zeldovich was often interested in understanding new phenomena, and had a habit of finding the best authorities in the new field to learn as much as possible from a collaboration with these people. Moscow is one of the best scientific centres in the world, and there are specialists in practically all fields. Quite often Zeldovich learned the basics of the new field very quickly and was able, either himself or in collaboration with specialists in the particular field, make significant contributions in the new field.
Why he selected our group for this task, I do not know. So far we had no experience in observational cosmology; our work was directed to the understanding of the structure of galaxies. We had theoretical cosmologists in our group (Enn Saar and Jaak Jaaniste), but they also did not have experience in observational cosmology. Thus, initially we had no idea how we can help Zeldovich.
But soon we remembered our previous experience in the study of galactic populations: kinematical and structural properties of populations remember their previous evolution and formation (Rootsmäe, 1961; Eggen et al., 1962). Random velocities of galaxies are of the order of several hundred km/s or less, thus during the whole lifetime of the Universe galaxies have moved from their place of origin only about1 h−1 Mpc (we use the Hubble constant in units H0 = 100 h kms−1 Mpc−1). In other words — if there exist some regularities in the large-scale distribution of galaxies, these regularities must reflect the conditions in the Universe during the formation of galaxies. Actually we already had some preliminary results: the study of companion galaxies had shown that dwarf galaxies are located almost s
olely around giant galaxies and form together with giant galaxies groups and clusters of galaxies. In other words — the formation of galaxies occurs in large units, not in isolation. A similar phenomenon is observed in star formation: stars born in star- forming gas clouds which evolve to form stellar associations, groups and clusters, but not as isolated objects (Ambartsumian, 1958).
Thus we had a leading idea on how to solve the problem of galaxy formation: We have to study the distribution of galaxies on large scales. Both our galactic astronomy and theoretical cosmology groups participated in this effort.
We started to collect redshift data from all available sources. Our first results from the study of the large-scale distribution of galaxies showed the existence of large superclusters in the Perseus as well as in the Coma and Hercules regions of sky (Einasto et al., 1975a). These superclusters contain several rich Abell clusters, numerous less rich Zwicky clusters, and ‘field’ galaxies. Both superclusters have a length over 50 h−1 Mpc, and form flat systems with axial ratio about 1:5. Thus the form of superclusters is far from a spherically symmetrical shape, and they have several rich clusters as nuclei. The main conclusion of the paper was: superclusters form from a gaseous medium prior to the formation of galaxies or simultaneously in a single process. By random clustering of galaxies it is impossible to form such dense and flat systems as superclusters. We noticed that individual galaxies also have a tendency to cluster at the same regions in space where clusters of galaxies are located.
When we started the galaxy distribution study, our Observatory had only two catalogues of galaxies, the de Vaucouleurs & de Vaucouleurs (1964) Reference catalogue of galaxies, and the Nilson (1973) Uppsala general catalogue of galaxies. In our studies of the nearby groups and dwarf satellite galaxies we made extensive use of these catalogues. The Uppsala catalogue is based on the Zwicky et al. (1968) Catalogue of galaxies and clusters of galaxies, so we needed the original Zwicky catalogues too.
