IAL 29: The Large-Scale Structure of the Universe

Don't Panic

Sections

  1. Introduction
  2. Units, Distance Scales, and the Observable Universe
  3. Groups, Clusters, Superclusters, and Large-Scale Structure
  4. Large-Scale Structure Formation and Evolution
  5. Galaxy Formation and Evolution
  6. Galaxy Formation and Evolution II



  1. Introduction

  2. The large-scale structure of the universe is the structure formed by galaxies (including absolutely postively their dark matter halos) treated as the smallest units---but one has to go to smaller scales too to understand how galaxies act and evolve.

    The large-scale structure includes galaxies, galaxy groups (see the figure of Seyfert's Sextet above: local link / general link: seyfert_sextet.html; and below: local link / general link: seyfert_sextet.html), galaxy clusters, galaxy superclusters, galaxy filaments, galaxy walls, voids, and whole lot more.

    The study of the large-scale structure is considered part of cosmology, but it is at a level just below that of the observable universe as a whole---which we will study in IAL 30: Cosmology.



  3. Units, Distance Scales, and the Observable Universe

  4. We need to orient ourselves a bit in the observable universe.

    There are some units and distance scales that it is convenient to note or recapitulate for a start. See the listing in the insert below (local link / general link: astronomical_distances_larger.html).

    Now what we can see of the
    observable universe (with observable universe radius = 14.3 Gpc) is a sphere centered on us.

    To explicate, see the artist's conception of the observable universe in the figure below (local link / general link: cosmos_artist_conception.html).

    To further explicate the observable universe, consider the 2 artist's conceptions of it in the figure below (local link / general link: observable_universe_cartoon.html).


    What is the
    observable universe made of? See the figure below (local link / general link: pie_chart_cosmic_energy.html).



  5. Groups, Clusters, Superclusters, and Large-Scale Structure

  6. Galaxies are NOT the largest structures.

    As we will discuss below, there are galaxy groups, galaxy clusters, galaxy superclusters, and the large-scale structure.

    Note that galaxy groups and galaxy clusters are probably mostly gravitationally bound: i.e., they will NOT expand with the expansion of the universe. On the other hand, galaxy superclusters though gravitationally interacting are probably mostly NOT gravitationally bound. In most cases, most of their component galaxy groups and clusters and field galaxies will move apart forever with the expansion of the universe---provided the Λ-CDM model or some other forever expanding cosmological models is true.

    Note velocity dispersion (σ) is a sort of average of the absolute values of the velocities of a set of astro-bodies (relative to their mutual center of mass) forming a gravitationally-bound system. The kinetic energy of the astro-bodies holds them up from collapse to the center of mass under their own self-gravity.

    1. Galaxy Groups:

      The structure just above galaxies themselves are galaxy groups.

      Characteristics of galaxy groups (see Wikipedia: Galaxy group: Characteristics):

      1. ∼ < 50 galaxies of Milky-Way luminosity (∼ 10**10 L_☉) or brighter (see Wikipedia: Galaxy group). There may be tens or hundreds of dwarf galaxies???---it is hard do a census of them since they are small and dim.
      2. Size scale ∼ 1--2 Mpc.
      3. Velocity dispersion (σ): ∼ 150 km/s.
        Recall, velocity dispersion (σ) is a sort of average of the absolute values of the velocities of a set of astro-bodies (relative to their mutual center of mass) forming a gravitationally-bound system.
      4. Mass ∼ 10**13 M_☉ of which ∼ 85 % is dark matter.
      5. Galaxy groups are the commonest grouping of galaxies in the local universe.
      6. ∼ 50 % of galaxies in the local universe are in galaxy groups.
      7. Galaxy groups are probably mostly gravitationally bound.???
      8. Some folks (like yours truly) just consider galaxy groups as VERY POOR galaxy clusters. It saves one from have to say galaxy groups and clusters all the time tediously.

      Our own Milky Way belongs to a galaxy group with the inspiring name of the Local Group.

      The Local Group is a very poor galaxy group: it has only 3 large galaxies.

      A map of the Local Group is shown in the figure below (local link / general link: local_group.html).


      Another well known galaxy group is the Hickson Compact Group HCG 87 which is shown in the figure below (local link / general link: galaxy_hcg_87.html).


    2. Galaxy Clusters:

      The structure just above galaxy group is galaxy cluster.

      The basic characteristics of galaxy clusters (see Wikipedia: Galaxy Cluster: Basic properties and Wikipedia: Galaxy Cluster: Composition):

      1. Total mass counting dark matter: ∼ 10**14--10**15 M_☉.
      2. Size scale: ∼ 1--5 Mpc.
      3. Velocity dispersion (σ): ∼ 1000 km/s.
        Recall, velocity dispersion (σ) is a sort of average of the absolute values of the velocities of a set of astro-bodies (relative to their mutual center of mass) forming a gravitationally-bound system.

      4. Their 3 main components:
        1. ∼ 50--few 1000s galaxies. But is this counting all dwarf galaxies????. Probably NOT, but it is hard to find reference to say this.???? The galaxies (including their dark matter) contribute only 1 % of the galaxy cluster mass (see Wikipedia: Intracluster medium: Composition).
        2. Intracluster medium (ICM): It is the hot plasma that permeates a galaxy cluster. The gas consists of ionized hydrogen (H) and helium (He) with approximately their primordial Big Bang nucleosynthesis (BBN) ratio of ∼ 73 % to 25 % plus a few percent metals (including iron (Fe)) ejected from galaxies by supernovae (SNe) and AGN feedback. The ICM is ∼ 10 % of the a galaxy cluster mass (see Wikipedia: Intracluster medium: Composition). The ICM is has temperatures in the range 10**7 to 10**8 K (see Wikipedia: Intracluster medium). The high temperatures is due too slow cooling plus heating by AGN feedback and/or shock heating from galaxy mergers (see Wikipedia: Intracluster medium: Heating).
        3. Dark matter: It contributes ∼ 84 -- 90 % of a galaxy cluster mass (see Wikipedia: Intracluster medium: Composition; Cimatti 2020, p. 165,174).

      5. There 3 types of galaxy clusters which are somewhat vaguely specified:
        1. Galaxy groups if you consider them as VERY POOR galaxy clusters. They contain ∼ < 50 galaxies of Milky-Way luminosity or brighter. Probably NOT counting dwarf galaxies????. See subsection Galaxy Groups above.
        2. POOR galaxy clusters: from galaxy group size to a few hundreds of galaxies. Probably NOT counting dwarf galaxies????.
        3. RICH galaxy clusters: from > POOR galaxy cluster size to thousands of galaxies. Probably NOT counting dwarf galaxies????.

        Galaxy clusters are also divided into REGULAR CLUSTERS which are roughly spherically symmetric with a concentration of galaxies in the center and IRREGULAR CLUSTERS which are irregular (FK-593).

        As usual, there is actually a continuum between the types of galaxy clusters.

        Also since the specifications are a bit vague, which category a galaxy cluster falls into will vary a bit with reference.

        For an example of a RICH galaxy cluster, see the Coma Cluster in the figure below (local link / general link: galaxy_cluster_coma.html).


      6. Probably, most galaxies are located in galaxy clusters (here counting galaxy groups as VERY POOR galaxy clusters???): see B. Ryden: Clusters and Superclusters.

        How many? Someone must have good statistics, but NOT yours truly. However, within 5 Mpc of the Milky Way about 80 % of galaxies are in galaxy groups and clusters (see Wikipedia: Field galaxies).

        POOR galaxy clusters far outnumber rich ones (FK-592) and most galaxies are NOT in rich clusters????.

        Galaxies NOT in galaxy groups and clusters (i.e., isolated galaxies) are called field galaxies.

      7. In fact, the number of galaxies in a galaxy cluster or galaxy group is often hard to determine since there are many dwarf galaxies (e.g., dwarf elliptical galaxies) that are hard to find (FK-594). In the Coma Cluster (see figure above: local link / general link: galaxy_cluster_coma.html), there are thousands for dwarf galaxies which we are only made obvious by imaging in the infrared.

        Also galaxy groups and galaxy clusters have NO sharp edges.

        This last comment applies both to the galaxies in the galaxy clusters and the dark matter which makes most of their mass.

      8. The galaxy clusters appear to be mostly gravitationally bound systems (FK-594).

          Question: Gravitationally bound means that the component objects of systems so described:

          1. CANNOT move indefinitely far apart without some extra kinetic energy. Their mutual gravitational attraction holds them in a group.
          2. are experiencing a mutual gravitational attraction.











          Both answers are right, but answer 1 is a more complete answer.

          Of course, complex interactions among objects in a gravitationally bound system may impart sufficient kinetic energy to individual objects to allow them to escape. In fact, on enough time-scale a chaotic gravitationally bound system will gradually disperse completely: objects chaotically gain enough kinetic energy to escape, leaving the others more tightly bound, but always with enough allow more escapes until all that is left is 2-body system which in Newtonian physics if completely isolated is eternally stable and repeating clockwork. The tendency of multi-body gravitationally bound systems to continuous escapes can be called "gravitational-system evaporation"

          Gravitational perturbations from outside the bound system may also lead to escape.

      9. The neareast RICH galaxy cluster is the Virgo Cluster.

        Now yours truly knows what you are thinking---"that's in Virgo"---well it straddles the line between the constellations Virgo and Coma Berenices.

        A sky map of constellation Virgo (with the Virgo Cluster shown) is shown in the figure below (local link / general link: iau_virgo.html).


      The relative locations of
      Local Group and the Virgo Cluster are shown in the figure below (local link / general link: large_scale_structure_030_mpc.html).


      Images of the central region of the
      Virgo Cluster in shown in the figure below (local link / general link: galaxy_cluster_virgo.html).


    3. Galaxy Superclusters:

      Galaxy clusters can themselves be part of superclusters that can contain tens of galaxy clusters and have a size scale of 50 Mpc (FK-594). Yours truly believes they are all irregular in shape.???

      Superclusters although gravitationally interacting appear to be mostly NOT gravitationally bound (FK-594). In most cases, most of their component galaxy groups and clusters and field galaxies will move apart forever with the expansion of the universe.

      We take up the expansion of the universe in IAL 30: Cosmology.

      The Local Group belongs to the Virgo Supercluster which itself is part of the Laniakea Supercluster.

      The Virgo Supercluster on the sky spreads out and reaches at least to the constellations Bootes, Coma Berenices, Leo, Ursa Major, and, of course, Virgo (see SLAS: Virgo Supercluster Overview Chart).

      The Virgo Supercluster, Laniakea Supercluster (sans name), and the local large-scale structure of the universe, is shown in the figure below (local link / general link: large_scale_structure_z_0x035.html).


    4. The Large-Scale Structure of the Universe:

      Galaxies and larger groupings collectively are called the large-scale structure of the universe.

      The larger structures include galaxy groups, galaxy clusters, galaxy superclusters, galaxy filaments (string-like bands of galaxies and larger groupings), galaxy walls (walls of galaxies and larger groupings), and voids (low density of galaxies).

      The voids are roughly spherical and have diameters of 30 to 120 Mpc (FK-596). The voids have some hydrogen gas and maybe??? filaments of dim galaxies.

      The study of the large-scale structure is actually considered part of cosmology.

      The 3-dimensional appearance of the large-scale structure of the universe is foamy (FK-596; CK-396). However, the modern tendancy is to think of it as web-like and hence the modern name for the large-scale structure the cosmic web.

      The local large-scale structure or cosmic web is illustrated in four figures below:

      1. Figure 1 of the local large-scale structure:

      2. Figure 2 of the local large-scale structure (see the figure below (local link / general link: local_universe_ir.html):


      3. Figure 3 of the local large-scale structure:

      4. Figure 4 of the local large-scale structure (see the figure below (local link / general link: large_scale_structure_z_0x035.html).


    5. Cosmological Principle:

      The cosmological principle is explicated in the figure below (local link / general link: observable_universe_cosmological_principle.html).


      The
      theory of the overall HOMOGENEITY and ISOTROPY of the observable universe is called the cosmological principle: the name, but NOT the concept, being introduced by E.A. Milne (1896--1950) (see figure below (local link / general link: e_a_milne.html).

      Of course, E.A. Milne (1896--1950) and other early users of the cosmological principle did NOT have modern evidence for it. For them, it was vastly simplifying assumption for research in cosmology.


      What
      homogeneity on a size scale of 400 Mpc means that every box of ∼ 400 Mpc on a side contains about the same amount of mass and more or less the same features on average.

      So there is a size limit to structure it seems: the cosmological principle scale ∼ 400 Mpc = 0.4 Gpc.

      So in a sense, we have to look over a considerable part of the observable universe to find the end of structure---the End of Greatness.

    6. No End of Greatness?

      It was once wondered if there was NO limit to structure---NO End of Greatness. That one would keep finding structure on larger scales.

      To explicate, say galaxy superclusters were grouped into Type II superclusters which were grouped into type III superclusters which were ... and so on forever. There would be NO End of Greatness and the mean mass-energy of the universe would go to zero if there were no mass-energy between galaxies. This cosmological model was considered once and was called the hierachical world model (Bo-14--15,19).

      A hierachical world model could even be a fractal.

      A fractal is a thing that is or looks the same on all or many scales at least in some approximation. The branches and roots of trees are approximate fractals.

      For example, one can make a fractal by drawing an iteration of 3 branches from each branch.

      The animation in the figure below (local link / general link: fractal_koch_snowflake.html) illustrates the construction of a fractal.

      But the observable universe is NOT a fractal nor any other kind of hierachical world model and does have a size limit on its structure. So there is End of Greatness: the cosmological principle scale ∼ 400 Mpc = 0.4 Gpc.




  7. Large-Scale Structure Formation and Evolution

  8. In Big Bang cosmology, the large-scale structure evolves from fluctuations in the density of the dark matter imprinted on the observable universe at a very early phase. The theory (or, more grandly, the paradigm) inflation (which we cover in IAL 30: Cosmology) gives a prediction for those fluctuations which so far has been adequate in that the given fluctuations are adequate initial conditions for (large-scale) structure formation and to explain fluctuations in the cosmic microwave background (CMB). The fluctuations in inflation were imprinted before ∼ 10**(-32) s after the Big Bang singularity of standard expanding universe models (which never happened in the view of most people).

    1. An Outline of the Story of Large-Scale Structure Formation: Reading Only:

      This section is for student reading only. It is best to discuss structure formation (i.e., the formation of the large scale structure of the observable universe or cosmic web) while explicating illustrative videos, and so we skip on to subsection Structure Formation in Videos below.

      So see the outline in the figure below (local link / general link: large_scale_structure_formation_outline.html).


      The figure below (
      local link / general link: large_scale_structure_early_universe.html) illustrates the large-scale structure of the universe in electromagnetic radiation (EMR) (NOT dark matter at cosmic time ∼ 4 Gyr (i.e., lookback time ∼ 10 Gyr).


    2. Structure Formation in Videos:

      Structure formation (i.e., the formation of the large scale structure of the observable universe is now discussed while explicating illustrative videos: first N-body simulation videos, second Large-Scale structure of the universe videos.

      For the videos, see also videos below (local link / general link: large_scale_structure_videos.html) further dynamically illustrate the formation of large-scale structure.

        EOF

        EOF

    3. Calculating Structure Formation:

      UNDER RECONSTRUCTION BELOW, BUT CAN BE READ FOR RECAPITULATION

      Structure formation (i.e., the formation and evolution of the large-scale structure of the universe) can only be done by computer simulations.

      The 1st order calculations are N-body simulations which include only dark matter which makes up ∼ 26 % of the mass-energy of the observable universe and ∼ 85 % of the CLUMPABLE mass-energy

      The dark energy makes up ∼ 70 % of the mass-energy of the observable universe, but it is thought NOT to clump.

      The N-body simulations only use gravitational forces.

      To go beyond the 1st order calculations, you need to include baryonic matter which is subject to gravitational forces, pressure forces, and magnetic force. There are also complex AGN feedback from central supermassive black holes.

      Computer simulation with baryonic matter are an ongoing and rapidly improving research area. A lot of success has been achieved in modeling the observable universe, but there is still a long way to complete understanding.

      ?????

      Computational simulations that start with a primordial distribution of density fluctuations and try to reproduce the modern large-scale structure are an ongoing enterprise with some success.

      But such computations are demanding and involve many approximations including the treatment of the dark matter and dark energy which we discuss briefly below in the section Galaxy Formation.

      The animation obtained from a computer simulation shown in the figure below (local link / general link: large_scale_structure_formation.html) dynamically illustrates the formation of large-scale structure.



  9. Galaxy Formation and Evolution
    SECTION UNDER RECONSTRUCTION BELOW, ALL A BIT REDUNDANT, A READING INSOFAR AS IT CAN BE READ

  10. Galaxy formation and evolution is a complex process and is NOT fully elucidated although we are learning more and more all the time from modeling and observations particularly of cosmologically remote regions which are timewise part of the early universe (FK-601--603).

    A major problem is the dark matter must have played major role in formation since it is overwhelmingly most of the mass. But that role is hard to determine exactly.

    What the dark energy does is another rather open question.

    1. Dark Matter and Dark Energy Redux:

      Dark matter is matter we only notice through its gravitational effect and it seems to be overwhelmingly most of the matter of the observable universe.

      About 20 % or so is baryonic dark matter. Most of this baryonic dark matter is hot and nearly invisible dilute hydrogen/helium gas between galaxies and galaxy clusters---the intergalactic medium which is mainly warm-hot intergalactic medium (WHIM)) and nearly-invisible intergalactic baryonic matter. Both these invisible matter types (which are mainly ionized hydrogen and helium with maybe a little enrichment of metals) radiate in the X-ray band, but with low detectability---we do observe it now a bit.

      However, when we (and everyone else) says dark matter without qualification, we (and everyone else) mean NOT the baryonic dark matter but exotic dark matter.

      The (exotic) dark matter may be some kind of very unreactive exotic particles that are spread through space, but clumped somehow in the dark matter halos of galaxies. People are trying to detect the exotic particles and there have been a few hints of detections, but nothing solid so far. On the other hand, the dark matter may be primordial black holes created in the Big Bang era before Big Bang nucleosynthesis (BBN).

      Dark energy is an unknown energy that seems to be accelerating the expansion of the universe. Dark energy is discussed in IAL 30: Cosmology.

      For the currently determined amounts of dark matter and dark energy (given that the Λ-CDM model is true), see the figure below (local link / general link: pie_chart_cosmic_energy.html).

      Both the concepts of dark matter and dark energy may be greatly modified or even dispensed with if the relativistic MOND theory turns out to real to some degree or other (Bekenstein, J. D. 2004, An Alternative to the Dark Matter Paradigm: Relativistic MOND Gravitation, astro-ph/0412652). See IAL 25: Black Holes for more on the relativistic MOND theory.


    2. The Small Large-Scale Structure of the Universe:

      We will only give a brief sketch of the current theory of galaxy formation: it is still a somewhat uncertain theory---and yours truly does NOT know as much as yours truly should know.

      The Big Bang (see IAL 30: Cosmology) left fluctuations in the density of the early universe gas as we mentioned above in connection with the large-scale structure.

      These fluctuations led to gravitational runaways to PROTOGALAXIES and these merged to make galaxies.

      The galaxies formed from the "the bottom up": (i.e., from smaller objects) rather from "top down" (i.e., out of galaxy-sized clouds) as was once theorized.

      The merger process never stopped: there are still mergers in the modern universe (i.e., the local universe).

    3. Galaxy Formation and Evolution: The Story So Far:

      UNDER RECONSTRUCTION BELOW

      Spirals (see figure below local link / general link: galaxy_hubble_sequence.html) formed from protogalaxies that were mainly still gas. The gas collapsed into disk according to the process we have discussed many times (see IAL 21: Star Formation). Then most stars formed in the disk.


      Computer simulations show that BAR FORMATION is quite natural for spiral galaxies. However, a sufficiently massive dark matter halo may inhibit BAR FORMATION---one theory anyway. The numbers of ordinary and
      barred spirals are comparable (CK-393; FK-583).

      It may be that SOME ellipticals formed when the early star formation was very fast and occurred before a disk of gas could form.

      There is then no energy dissipation mechanism for stars which are point-like and super-rarely collide in a hard sense: gravitational interactions are common but these tend NOT to dissipate kinetic and gravitational potential energy to heat.

      Thus the streaming and dissipation mechanism that leads to disk formation for gas does NOT occur for stars.

      Consequently, the stars in ellipticals stayed in a swarm.

      However, very likely most ellipticals form by galaxy mergers. Maybe they almost all do.

      The process of galaxy merger randomizes the orbits of the stars and often strips of the gas.

      The stars go right through each other in galaxy collisions, but the gas runs into gas clumps and becomes physically separated and tends to be lost which tends to turn off star formation. Or the gas becomes too hot for star formation. Its pressure is too high.

      For a galaxy collisions see the video When galaxies collide! | 1:36 in Galaxy videos below (local link / general link: galaxy_videos.html):

        EOF

      The
      merged galaxy will strong tend to be an elliptical without much star formation because without much gas or at least much cold enough gas.

      The merged galaxy (likely a new elliptical) tends to be quenched galaxy in modern astro-jargon.

      Why do quenched galaxies NOT revive by inflow of new gas and gas cooling?

      They probably do or did in the past in some cases.

      But it seems that dequenching is rare.

      The fact that a galaxy merger occurred probably means crowded ????

      Also star formation is slowing down generally as the universe expands and there is less gas inflow from the intergalactic medium (IGM).

      UNDER RECONSTRUCTION BELOW

      The distinction in formation between spirals and ellipticals probably has to do with the richness of the environment.

      Ellipticals are more likely in regions of greater density of galaxies.

      Ellipticals can also be made by the collisions of spirals which strip the gas and randomize the star orbits and strip away the interstellar medium (ISM).

      in the figure below (local link / general link: galaxy_mice.html).


      Mergers have evidently happened in rich clusters: distant rich clusters have more spirals than local ones.

        Question: Distant rich clusters are rich clusters seen in an earlier epoch of the universe because of:

        1. finite travel speed of light.
        2. infinite travel speed of light.
        3. the universal expansion.










        Answer 1 is right.

        In the universe, to look far away is to look long ago.

      Collisions of
      ellipticals in dense environments probably also strip any new gas from ejecta from old stars through winds. This mostly prevents renewal of star formation.

      See the figure below for further discussion of ellipticals, galaxy mergers, and galaxy quenching (the turning of star formation).


      Rich clusters typically have a very hot intergalactic medium that radiates X-rays: temperatures of order 10**7 and 10**8 K.

      Such gas is probably shocked heated during collisions that stripped it from the parent galaxies (FK-596).

      In both ellipticals and spirals the star formation rate probably peaked early on in the universe.

      The adjacent cartoon illustrates this, but the cartoon simplifies things by ignoring that some spirals have merged to make ellipticals.


  11. Galaxy Formation and Evolution II

  12. UNDER RECONSTRUCTION BELOW: DO NOT READ How do Ellipticals Form---and for that Matter Spirals?

    Well let's go into the stories with lots of simplifications and omissions to keep from turning our discussion into a book.

    1. Early in cosmic time (e.g., of order 10--14 Gyr ago), dark halos fromed from density perturbations in the primordial distribution of dark matter.

      The "rich get richer, the poor get poorer".

      Gravitational runaway led to dark halos where the dark matter where the density was high and voids where it was low.

      The dark halos formed gravitational wells that attracted ordinary matter almost entirely the hydrogen and helium in a 3:1 ratio.

      If star formation proceeded very rapidly in a dark halo due to some initial condition????, most of the gas is used up to form stars in a few early generations of stars leaving little for later generations of stars.

      Also there was no time for the gas to relax to galactic disk which requires pressure interactions which stars do NOT have---they pinpricks that interact gravitationally virtually only.

      elliptical galaxy.

      If there are no interactions with other galaxies, the elliptical galaxy only ages through the aging of its stars.

      As cosmic time passes, progressively the less massive stars end their main-sequence lifetimes and turn into white dwarfs.

      The end of the main-sequence lifetime must inject some matter back into the interstellar medium (ISM), but evidently NOT enough to promote much new star formation.

    2. What if star formation proceeds more slowly through cosmic time from the clumping of ordinary matter in a dark halo?

      Stars keep forming up to the present in cosmic time and far into the future.

      The ISM has time to relax through pressure interactions into a galactic disk where almost all the star formation happens.

      Usually there are spiral arms for reasons that we go into section spiral arms.

      The result is a spiral galaxy.

    3. What if you have galaxy mergers?

      These are, in fact, common.

      Galaxy mergers are complex events that tend to strip ISM and randomize star orbits of merging spiral galaxies or merging spiral galaxies elliptical galaxies.

      So one tends to end up with elliptical galaxies.

      Merging elliptical galaxies just give larger elliptical galaxies.

      Both calculations and empirical evidence show that mergers yield elliptical galaxies.

    4. So elliptical galaxies can be born elliptical galaxies or come from galaxy mergers.

    5. Galaxy mergers are most likely in galaxy clusters.

      The more compact the galaxy cluster, the more galaxy mergers will occur. So NOT surprisingly that compact galaxy clusters tend to have a higher fraction of ellipticals than less compact galaxy clusters.

    6. The larger ellipticals (which are the largest galaxies overall) are often found at the centers of rich (i.e., large) galaxy clusters.

      These are the cD or giant elliptical galaxies: e.g., M87 in the Virgo Cluster in Virgo.

      The growth of massive ellipticals is called galactic cannibalism---they eat their own.

      This growth is illustrated in the figure below.