Also we needed data on peculiar galaxies, since our previous experience has shown that active galaxies are good markers of the skeleton of systems of galaxies — they mark the locations of dense regions of space where ordinary galaxies also have a tendency of clustering. We already had the lists of Markarian galaxies andVorontsov-Veljaminov interacting galaxies; now we needed the Zwicky Catalogue of Selected Compact Galaxies (Zwicky & Zwicky, 1971), as well as Zwicky catalogues of galaxies and clusters of galaxies. Our Observatory had no money to buy it. So I wrote a letter to Fritz Zwicky asking advice on how to get the catalogues. In his answer Zwicky (see Fig. 5.2) explained that the Compact Galaxies catalogue was published privately, and in order to get it we had to pay 205 Swiss francs to his daughter who lives in Switzerland. Now the problem was how to get the money.
Fig. 5.2 Letter by F. Zwicky concerning the purchase of his Catalogue (author’s archive).
During the IAU General Assembly in Brighton 1970 I met Tarmo Oja, an Uppsala astronomer of Estonian origin. We spent a lot of time together during excursions; I have a photo with him and his wife in the new Greenwich Observatory in Herstmonceux Castle near Brighton. Next time we met was in Athens during the First European Astronomy Meeting. We started to exchange literature — Tarmo subscribed for me “Astronomy and Astrophysics” (Tartu Observatory did not have money for this), and I sent Tarmo Estonian books. Based on this experience I wrote a letter to Tarmo asking whether he could help us and order the Zwicky catalogues for me. He answered very quickly and wrote me that the money had been sent to the address given by Zwicky, and for the Catalogue of galaxies and clusters of galaxies to the California Institue of Technology. All catalogues arrived rather soon.
The Uppsala catalogue contains all data on Zwicky galaxies, but not on Zwicky clusters. With these catalogues we were able to study the 2-dimensional distribution of galaxies of various type in the sky, as well the distribution of groups and clusters of galaxies. Of special interest were Zwicky clusters of distance class “near”, since it was the nearby space from where we hoped to get enough redshift data to find the spatial 3-dimensional distribution of all astronomically interesting objects from normal and active galaxies to groups and clusters of galaxies. The Estonian Academy of Sciences was able to buy for the Tartu Observatory the Palomar Atlas — paper prints of all photos of the Palomar Observatory Sky Survey made with the large Schmidt telescope. This allowed us to see the structures we found in catalogues as they appear in reality.
A few years later we had the possibility of getting the de Vaucouleurs et al. (1976) Second Reference catalogue of galaxies. We also used a number of other sources such as the Sandage (1978) list of redshifts for 719 bright galaxies, and the Sandage & Tammann (1981) Revised Shapley–Ames catalog of bright galaxies. An additional source was the ZCAT — a compilation of all available galaxy redshifts by John Huchra.
In our work to solve the Zeldovich question we had a close collaboration with his team. In 1975 Sergei Shandarin obtained first results of numerical simulations of the evolution of particles according to the theory of gravitational clustering developed by Zeldovich (1970), see Fig. 5.3. Results of the simulation were published a few years later by Doroshkevich & Shandarin (1978). A figure with the results of these simulations was put on the wall of the Saar’s and Jaaniste’s office. In this picture a system of high- and low-density regions was seen: high-density regions form compact clumps and are joined by filaments, together they form a cellular network which surrounds large under-dense regions. This simulation was, however, only 2-dimensional and had a qualitative character — we did not know the scale of the network seen in the simulation.
Our challenge was to find out whether the real distribution of galaxies shows some similarity with the theoretical picture. We wanted to understand the actual distribution of galaxies and systems of galaxies, and to compare the real distribution with model distributions of all theoretical scenarios. We did this, step by step, and our final conclusion was that none of the theoretical scenarios suggested so far meets all observational constraints. But this was a long way. Our immediate goal was to compare the actual distribution with the prediction in hand — the Zeldovich pancake scenario.
Fig. 5.3 Distribution of particles in simulations (Shandarin 1975, private communication), Fig. 7 of Doroshkevich & Shandarin (1978) and Fig. 6 of Einasto et al. (1980a)).
If we identify rich knots in the simulated structure with superclusters, then results of simulations suggest that superclusters must be joined by galaxy or cluster chains to a connected network. To identify these chains it is needed to study the global distribution of galaxies and clusters on large scales, not just the local environment of superclusters. The Peebles scenario of galaxy formation predicted a more-or-less random distribution of galaxies; the Ozernoi scenario did not have any prediction on the distribution of galaxies. Thus we hoped that the large-scale distribution of galaxies and systems of galaxies contains the key information to discriminate between basic scenarios of galaxy and structure formation.
To find the global distribution of galaxies and systems of galaxies we used several methods. Jaaniste and Saar suggested studying the distribution of nearby Zwicky clusters. First of all, the Zwicky catalogue on near clusters is complete,and many bright galaxies of nearby Zwicky clusters had at this time measured redshifts, so we hoped to determine the distribution of clusters, and to find some regularities there. To see the distribution better we built in the office of Saar and Jaaniste a 3-dimensional model from plastic balls. Some regularity was evident: there were several clusters of Zwicky clusters — superclusters — one of them in the Perseus region. But too many clusters had no galaxies with measured redshifts, so it was difficult to get an overall picture.
Mihkel Jõeveer found a simple method how to estimate mean redshifts of all near clusters. Most near clusters contained enough member-galaxies to find the cluster luminosity function using apparent magnitudes. By comparing these functions with similar functions for clusters with known redshift it is easy to estimate the redshifts of all clusters. The sample of Zwicky near clusters was used in our preprint by Jõeveer et al. (1977), and by Einasto et al. (1980a) (Figs. 5.7, 5.8 and 5.9) and by Tago et al. (1984, 1986).
Our young collaborator Erik Tago started to search nearby systems of galaxies using prints of the Palomar Sky Survey. His first goal was to look for nearby galaxy systems in the Coma and the Virgo supercluster region. We all followed his search with curiosity. Soon a poor cluster was found, then the next one. Later, when Zwicky catalogues of galaxies were available, we discovered that all these clusters Erik had found are listed as Zwicky’near’ clusters. Initially there were no redshifts for these galaxies. As our redshift compilation improved Erik was able to find distances to these clusters. The clusters formed a chain though a large void between the Virgo, the Coma and the Hercules superclusters. Tago et al. (1986) argued that this string of galaxies can be an evidence for a Lagrangian Singularity in the Zeldovich pancake scenario.
A third approach was applied by Mihkel Jõeveer. He used wedge-diagrams, invented just when we started our study. He applied two ideas to understand the spatial structure of the distribution of galaxies and systems of galaxies. Most important was his idea to investigate the distribution not only of ordinary galaxies, but all other interesting astronomical objects — peculiar galaxies, such as Markarian galaxies, radio galaxies, galaxies from the Zwicky list of compact galaxies, and of various systems of galaxies from groups to clusters. His second idea was: he made a number of relatively thin wedge diagrams in sequence, and plotted in the same diagram ordinary and active galaxies, as well as groups and clusters of galaxies. By comparing neighbouring wedges in declination and in right ascension it is possible to understand the three-dimensional distribution of objects.
In these diagrams a regularity was clearly seen: galaxies and galaxy systems populate identical regions, and the space between these regions is empty. Ordinary galaxies are mostly located along a network of
strings or filaments. Groups and clusters of galaxies, as well as active (Markarian and radio galaxies) form the skeleton of the network. Galaxies lie close to the skeleton, or form bridges between elements of the skeleton. After this success we concentrated our efforts to the study using wedge-diagrams. Most attention was directed to the Perseus–Pisces supercluster, well seen also in our model with balls as a cluster of Zwicky clusters. In this region there are almost no foreground galaxies as shown already by Rubin et al. (1973), thus we see the structure of the supercluster very clearly.
When we had our first results using both Zwicky clusters and wedge diagrams with all objects, we immediately understood that distribution of galaxies and of systems of galaxies tells us something principally new about the formation and evolution of the Universe — the Universe has structure, it is not structureless as believed so far. We had again the feeling that we have reached a tip of a mountain, and behind the peak there is a completely new and unexplored landscape. Also we understood that whatever we did next, we would be the first to do it — it was a completely new area of research.
To our great surprise and joy slices with galaxies and clusters were quite similar to the theoretical picture predicted by Zeldovich. We made wedge-diagrams for the full sky, for three declination wedges they are shown in Figs. 5.4, 5.5 and 5.6. Wedge diagrams were shown in the Tallinn symposium, Fig. 5.4 was published in the symposium proceedings by Jõeveer & Einasto (1978). We also made a very detailed analysis of the Perseus supercluster region and made pictures of the distribution of Zwicky clusters of the class “near” in the Perseus region of sky, see Figs. 5.7, 5.8 and 5.9 for clusters in three redshift intervals. The distribution of Zwicky clusters in the Perseus region was shown in our preprint by Jõeveer et al. (1977), and published in detail by Einasto et al. (1980a). Similar distributions were found for the Coma region of sky, published a few years later by Tago et al. (1984).
5.2.2 The Tallinn symposium on large scale structure of the Universe
Already in 1975, after the Tbilisi Meeting, we discussed with Zeldovich the possibility of organising a real international conference devoted solely to cosmology. Due to the Soviet bureaucratic system it was extremely difficult for Soviet astronomers to attend international conferences in Western countries; thus the only possibility of having better contact between Soviet and Western cosmologists was to hold the conference within the Soviet Union or in nearby “Socialist” countries. Actually a conference on cosmology was recently held in Krakow, Poland, during the IAU General Assembly, where Zeldovich had a talk on his scenario of structure formation (Doroshkevich et al., 1974).
Fig. 5.4 Wedge diagram for the 30°–45° declination zone. Filled circles show rich clusters of galaxies, open circles — groups, dots — galaxies, crosses — Markarian galaxies. Clusters at RA about 2 h belong to the main chain of clusters and galaxies of the Perseus–Pisces supercluster; galaxies and clusters near the center at RA about 12 h are part of the Local supercluster, and galaxies and clusters at redshift about 7,000 km/s and RA between 10 h and 13 h belong to the Coma supercluster. Note the complete absence of galaxies in front of the Perseus–Pisces supercluster, and galaxy chains leading from the Local supercluster towards the Coma supercluster (Jõeveer & Einasto, 1978).
Zeldovich suggested holding the symposium this time in Estonia — here the organisation was much more easier than in Moscow. The only suitable place was our capital Tallinn, which had conference facilities, and was open to foreign vistors, whereas Tartu was at this time a closed city due to its proximity to a large military airfield. Initially I suggested having a discussion on dark matter, but in Zeldovich’s opinion this topic was still too controversial, and it would be better to have a more neutral topic. After some discussion we decided to devote it to “Large Scale Structure of the Universe”. At this time we had no idea what this term actually means, I had one example of a planned IAU symposium on “Large Scale Characteristics of the Galaxy”, to be held in 1978 in College Park, USA.
Fig. 5.5 Wedge diagram for the declination zone 15°–30°. Galaxies and clusters near the center at RA about 12 h belong to the Local supercluster; two rich clusters at redshift 7,000 km/s and RA about 12 h are the main clusters of the Coma supercluster, galaxies at redshift 5,000 and RA about 23 h belong to the Perseus–Pisces supercluster. Note the galaxy chains leading from the Local supercluster towards the Coma and the Perseus–Pisces superclusters (Jõeveer & Einasto, 1978).
To discuss the program of the planned symposium the scientific organising committee met during the IAU General Assembly in Grenoble 1976. I was the sole representative from the Soviet Union, and we discussed both the program and the candidate for the chairman of the SOC. Most members of SOC were from Western countries and wanted the chairman to also be from a Western country. So I suggested Malcolm Longair. He had spent a long period in Moscow and was well familiar with the work of Moscow theorists, and Zeldovich fully trusted him. This suggestion was accepted and, as time has shown, was very good — he was really a very effective leader.
Fig. 5.6 Wedge diagram for the declination zone 0°–15°. Galaxies and groups at redshift 5,000 between RA 23 h and 1 h belong to a supercluster with Abell cluster A194 as the main cluster, see Fig. 5.7. Clusters at redshift about 13,000 km/s at RA between 23 h and 1 h belong to a system of superclusters at the far side of a large void, seen in Figs. 5.7 and 5.9. Note the galaxy chains leading from the Local supercluster towards the A194 supercluster (Jõeveer & Einasto, 1978).
In preparations for the symposium we had two goals in mind. The first one was scientific. In 1976 we already had preliminary results of our own study of the distribution of galaxies and clusters, and knew that only a discussion of global properties of the distribution can answer basic theoretical questions concerning the formation and evolution of the Universe. Thus we suggested discussing all major global aspects of the observational and theoretical cosmology. At first glance this seems to be too broad a scope, but, as results of the symposium showed, this broad scope was needed to understand major theoretical problems on the evolution of the Universe.
Fig. 5.7 The distribution of Zwicky clusters of the distance class ‘near’ in the Perseus area of the sky in redshift interval 3, 500 ≤ V0 < 6, 500 km/s. Abell clusters A426, A347, A262 and Zwicky clusters 37, 31, 20, 10, 6, 5 form the main chain of clusters of the Perseus–Pisces supercluster. The Northern chain is formed by Zwicky clusters 30, 22, 498, 497, the Southern chain by Zwicky clusters 40, 38, 33, 35, 29, 24; this chain forms a bridge towards the supercluster which surrounds the A194 cluster (Jõeveer et al., 1977; Einasto et al., 1980b).
Our second goal was of a completely different nature. I had attended previously several international astronomical meetings elsewhere in the Soviet Union, and had been astonished by the rather bad organisation. Soviet scientists wanted Russian to be a major scientific language, and quite often talks by Soviet astronomers were in Russian with no adequate English translation. Also it was difficult to follow discussions because of the language conflict. Other aspects of local organisation were also poor — meals, accommodation, cultural programs. Thus we had a very special goal — to prepare the symposium on a really high international level.
First of all, we assembled our Technical Organising Committee (TOC) using almost all active astronomers of Tartu Observatory, not only of our cosmology team. This Committee worked completely independently of the Scientific Organising Committee, and had full freedom and possibilities for any initiative. Among us we called the TOC “Glavsympstroy” to parody the Soviet style abbreviations of long names of organisations. There were people responsible for auditorium, for equipment, for distribution of the guests to hotels, for the cultural program, for English translation etc. Some of the TOC members had good relations with the Moscow circles, so we were able to hire the best synchronous translators from Moscow, really good professionals. This allowed us to have all talks and discussions in English; the translation was so good that even people with bad understanding o
f English could follow talks and discussions.
Fig. 5.8 The distribution of Zwicky clusters of the distance class ‘near’ in the Perseus area of the sky in redshift interval 6, 500 ≤ V0 < 10, 000 km/s. Clusters A397, A400 and numerous Zwicky clusters form a cluster system at R.A. α ∼ 3h, clusters A2634, A2666 and surrounding Zwicky clusters form another cluster system at R.A. α ∼ 23h. These cluster systems surround a large void between them, the Perseus–Pisces supercluster in the foreground, and another supercluster in the background, seen in the next redshift interval (Jõeveer et al., 1977; Einasto et al., 1980b).
Some members of the TOC knew Estonian people active in music, so it was possible to organise for the symposium a special concert in the historical Tallinn Rathaus (Town Hall) where a young conductor Tõnu Kaljuste gave a concert of classical music conducting the recently formed Chamber Choir. In the following years this choir has given concerts all over the world and is counted as one of the very best. In 1977 the choir was not yet internationally known, and astronomers were surprised to hear a concert at this high level.
Fig. 5.9 The distribution of Zwicky clusters of the distance class ‘near’ in the Perseus area of the sky in redshift interval 10, 000 ≤ V0 < 15, 000 km/s. Numerous Abell and Zwicky clusters form in this region several rich superclusters. Together with superclusters seen in lower redshift regions these superclusters surround a large void (Jõeveer et al., 1977; Einasto et al., 1980b).
The symposium was held in September 1977 with reception, cultural program, and banquet as normal for an international meeting. All talks and discussions were tape-recorded, and discussion sheets edited and typed by our secretaries, who sometimes worked the full night to prepare the typed text for further editing. This allowed Malcolm to take with him the full set of typed discussions. To help final editing a copy of all tapes was given to Malcolm. All this helped to prepare the symposium proceedings. This symposium was of such interest to the Russian speaking community that a full Russian translation was soon prepared and printed.
Dark Matter and Cosmic Web Story Page 17