Friday, February 19, 2010

Astronomy- Vesto Slipher and Galaxy Redshifts

This short essay describes the work of Vesto Slipher, a rather unsung astronomer who at the turn of the 20th century made many of the fundamental observations that led to Edwin Hubble's discovery of an expanding universe.

Vesto Melvin Slipher – Astronomical Hero
W. Keith Fisher

I. Introduction

Vesto Melvin Slipher was born in Mulberry, Indiana in 1875 and educated in Astronomy at the University of Indiana, where he received his doctorate in 1909. Although Slipher received many honors during his scientifically productive years as the world’s foremost astronomical spectroscopist, he was overshadowed in his most important contribution to science, the measurement of radial velocities of spiral nebulae (now known as spiral galaxies), by the flamboyant Edwin Hubble (General Notes 1969). Slipher being a humble, modest man never demanded his share of the glory. In this essay I will first discuss Slipher’s contributions that make him an astronomical hero then propose some reasons why his name and fame have faded over time.

In 1900 Percival Lowell ordered a custom-built spectrograph built by the John A Brashear, Co. to couple to the 24 inch Clark refractor at his observatory (appropriately named the Lowell Observatory) in Flagstaff Arizona. In 1901 he hired Slipher, who arrived at Flagstaff that summer (Bartusiak 2009).

It turned out that Lowell chose the right man. During the first two decades of the twentieth century Slipher made many important contributions to astronomy through spectroscopy. He took this spectrograph, intended for planetary work, and with great skill modified it to measure spectra of galaxies. As we will see Slipher found himself confronting results that gave the first indication that we live in an expanding universe (Bartusiak 2009; Slipher 1917).

II. Sliphers Accomplishments

When Slipher first arrived at Lowell his emphasis was on developing his expertise in using the 24” refractor plus spectroscope, which he did very successfully, eventually becoming a world leading expert on planetary spectroscopy. He was painstaking in his attention to detail regarding calibration and precise measurement of the wavelength of spectral bands (Slipher 1917). This emphasis on planetary spectroscopy was not surprising since his financial support came from Percival Lowell, whose interest was in planetary science, especially observations of Mars in his quest for canals.

In 1903 Slipher undertook a spectroscopic investigation of the rotational period of Venus. In this work he developed the inclined line method of measuring rotational period. For this measurement the slit of the spectroscope was oriented perpendicular to the rotation axis of Venus so that part of the light coming through the spectroscope slit was from the approaching side of Venus while the remainder was from the receding side. The result was that the spectral lines were inclined, and with the proper mathematics the velocity of rotation could be derived. He went on to use this method to measure the rotation rate of Mars, Jupiter, Saturn and Uranus (Slipher 1903; Slipher 1903a).

Slipher followed up this work with a photographic study of the spectra of Neptune and Uranus. Exposure times were long, 14 hours and 21 hours respectively. Using his careful calibration techniques he was able to determine the wavelength of observed spectral bands with good accuracy (Slipher 1904).

Then in 1905 and 1906 he continued working toward his goal to determine the condition and substance existing in the atmospheres of planets, with photographic studies of Jupiter and Saturn. After experimentation with red sensitive photographic plates, He obtained photographs of Jupiter and Saturn in the red-orange-yellow spectral region (Slipher 1905; Slipher 1906a).

During 1906 and 1907 he did extensive experimentation on methods to sensitize the plates to red light. He developed a method of stretching the spectral response of plates all the way out to 700 nm. He used plates sensitized in this manner to again photograph the spectra of Jupiter, Saturn, Uranus and Neptune. Many spectral lines were identified and differences in the spectra of these planets noted in detail but almost nothing was concluded regarding the chemical identity of the bands. Hydrogen was identified in all four planets with perhaps a low concentration of water vapor (Slipher 1906). In later work some of these lines were identified by Slipher and others to be from ammonia and methane (Abrahams).

In addition to his groundbreaking work on planetary spectroscopy he found that nebulae such as the Merope nebula in the Pleiades had the same spectral signature as that of the other nearby bright stars. He concluded that the nebula was “pulverized matter” that was shining by reflected light. Slipher had just discovered reflection nebula (Lowell & Slipher 1914). In addition he discovered interstellar sodium and found the first evidence of interstellar calcium, when in the spectrum of a binary system, the calcium line showed no oscillation (Bartusiak 2009; Abrahams). Slipher studied the radial velocities of globular clusters, the spectra of comets and the aurora. However it was his spectroscopic studies of spiral nebulae, the work where he failed to garner his fair share of credit that was his most significant contribution to astronomy.

His studies on spiral galaxies were motivated by a letter from Percival Lowell in 1909 asking him to use his red sensitive plates to photograph the spectra of spiral galaxies. At first Slipher balked at this request. Because of the slow speeds of photographic emulsions in 1909 he knew that the exposure times would be at least 30 hours with the long focus refractor. This telescope did not have sophisticated tracking capability, so keeping the slit of the spectrograph trained on a spiral nebula over the long exposures would be a formidable task. Slipher cleverly modified the spectrograph to operate 200 times faster than the original specifications and he carried on with the work (Bartusiak 2009).

The first glimmer of success came on September 17, 1912 when he was able to photograph a spectrum of the Andromeda Galaxy using an exposure time of nearly 7 hours. He showed for the first time that the Andromeda Nebula was approaching the Sun with the extraordinary velocity of 300 km/s, more than ten times the rate of average stars. He also measured its rotation rate using the inclined line technique and found it to be very high. Further work though 1912-1913 showed that spiral nebula as a class have a much higher order of velocity than have the stars (Lowell & Slipher 1914; Slipher 1917).

Of the twenty five radial velocities measured, only four were approaching, the remainder were receding and at extremely high velocity. In 1917 Slipher did not appreciate the meaning of the preponderance of receding spirals. His hypothesis at the time was that they were indicative of motion of our stellar system relative to the other spiral nebula. He did realize that, based on the star-like characteristics of their spectra, the spirals contained stars. But it was the high radial velocities of the spirals that led him to believe that they were an entirely different class of objects from stars, globular clusters and diffuse nebula. His observations eventually led him to embrace the “island universe” theory, which regards our stellar system and the Milky Way as a great spiral nebula seen from within and that the spiral nebula were at great distance relative to the stars (Slipher 1917).

Slipher presented his work on radial velocities of spirals at Northwestern University in 1914. In the audience was an ambitious young astronomer, Edwin Hubble, who realized that the only way to confirm the island universe theory was to measure the distances of these spirals. This occurred in 1923-24 when Hubble, using the 100 inch telescope on Mt. Wilson, identified Cepheid variables within the Andromeda Nebula and used their pulsation period and their period-luminosity relation to establish that it was indeed at very great distance and was a separate island universe. About five years later, working with Milton Humason, Hubble identified the proportionality between radial velocity and distance, the constant of proportionality became know as the Hubble constant (Hubble 1929).

III. Why Slipher Isn’t Famous

At issue here is that Hubble failed to give Slipher the credit he deserved for the radial velocity data in his famous 1929 paper where the relation between distance and radial velocity was first reported. To put in perspective the magnitude of this nefarious deed, by 1925 Slipher had measured the radial velocity of about 40 spiral nebula while only four radial velocity measurements had been done by astronomers in systems that Slipher had not studied first (Gribbin 2002). It was not until a lecture given in 1953 did Hubble give credit where credit was due. In this lecture he professed that his discovery of the distance-velocity relation in 1929, which was later realized to be indicative of an expanding universe, “emerged from a combination of radial velocities measured by Slipher at Flagstaff with distances derived at Mt. Wilson” (Bartusiak 2009).

So Hubble’s failure to recognize Slipher's accomplishment is one reason for his relative obscurity among astronomy’s elite, but it is not the only reason. Another reason is Slipher chose to publish most of his results in brief accounts in the Lowell Observatory Bulletin instead of a major astronomical journal (Bartusiak 2009). He did publicize his radial velocity measurements more widely but he was steamrolled by Hubble.

A third reason is that Slipher did very little original astronomical work after 1933. He became more interested in business pursuits and financial security, which was quite understandable as he was living through the depths of the Great Depression. However he remained director of the Lowell Observatory and let the facility slide into stagnation. It was not until the 1950’s, after Slipher resigned, that the Lowell Observatory revamped and was brought into the twentieth century (Tenn 2007).

So that’s it, Hubble gets a fundamental physical law named after him, credit for the discovery of the expansion of the universe and a large space telescope in his name. Slipher gets some belated credit. But it should be realized that, despite his unproductive later years, his measurements provided half of the critical data set that led to the discovery of the expansion of the universe and that his methods and the equipment he developed during the first two decades of the twentieth century were at the very cutting edge of technology for its time.


Abrahams, Peter, “Early Instruments of Astronomical Spectroscopy,” web site
Bartusiak, Marcia, 2009, Sky and Telescope, Vol. 118, No. 3, p. 30-35.
Hubble, Edwin, 1929, “A Relation Between Distance and Radial Velocity Among Extra-Galactic Nebulae,” in The Early Universe: Reprints, ed. E. W. Kolb and M. S Turner, (Redwood City, CA., Addison-Wesley Publishing Company), p. 9.
General Notes, 1969, Publications of the Astronomical Society of the Pacific, Vol. 81, No. 483, p. 922.
Gribbin, John, 2002, The Scientists, Radom House, New York, 591.
Lowell, Percival and V.M. Slipher, 1914, “Epitome of Results at the Lowell Observatory, April 1913 to April 1914,” Lowell Observatory Bulletin, Vol. II, No. 59.
Slipher, V.M., 1903, Lowell Observatory Bulletin, Vol. 1, No. 3.
Slipher, V.M., 1903a, Lowell Observatory Bulletin, Vol. 1, No. 4.
Slipher, V.M., 1904, Lowell Observatory Bulletin, Vol. 1, No. 13.
Slipher, V.M., 1905, Lowell Observatory Bulletin, Vol. 1, No. 16.
Slipher, V.M, 1906, Lowell Observatory Bulletin, Vol. 1, No. 42.
Slipher, V.M, 1906a, Lowell Observatory Bulletin, Vol. 1, No 27.
Slipher, V.M., 1917, “Nebulae,” Proceedings of the American Philosophical Society, Vol. 57, p. 403-409. Reference provided by the NASA Astrophysics Data System.
Tenn, Joseph S., 2007m Journal of Astronomical History and Heritage, Vol. 10, No.1, 65.

Astronomy - The Red Shift Controversy

I have an intense interest in Astronomy. It is the field of science that attempts to provide rationl answers to the big question such as the origin of the universe. Below is a paper I wrote regarding some controversy about some fundamental observational underpinnings of Big Bang Cosmology. The controversy centers around the question if observed red shifts of galaxies is indicative of their velocity of recession. This controversy has largely faded into history because there is overwelming evidence that red shifts do indicate recessional velocity.

The Red Shift Controversy
W. Keith Fisher

I. Introduction

The red shift controversy had its origins in the mid 1960’s when Halton Arp observed that certain apparently physically associated object had widely differing redshifts. His initial observations concentrated on associations of large redshift quasars with low redshift active galaxies. He proposed that the large redshifts observed for quasars were non-velocity, non-cosmological in nature and that they were intrinsic to the quasars themselves. No physical mechanism was suggested for these intrinsic redshift other than they were caused by some new physical process that occurs in quasars. Further observations showed that some high redshift quasars look to be physically associated with low redshift galaxies, that redshifts seem to have certain preferred values, that associations of galaxies also had disparate redshifts but to a lesser extent that quasar-galaxy pairs, and that the distribution of quasars over the sky was not isotropic.

Today the vast majority of astronomers discount Arp’s intrinsic non-velocity related redshift hypothesis. Most astronomers question the statistical methods Arp used to determine if objects of discordant redshift were really associated, or just chance alignments. It is the position of the majority of the astronomical community that the conventional interpretation of redshift, as velocity related and indicative of distance in an expanding universe, provides the most natural and self consistent explanation within the framework of the Big Bang Theory. However Arp and a few other astronomers still promote intrinsic redshifts as at least a component of the total observed redshifts for quasars and active galaxies. Resolution of this controversy is extremely important to astronomy and science in general because, if the non-cosmological interpretation turns out to have merit, it will require a major revision of our current understanding of cosmology and many of the physical principals we use to describe the origin and evolution of the Universe.

A good place to start our discussion is with the conventional view of quasars. Quasars are characterized by very large redshifts of their emission lines. The conventional explanation of these redshifts is a cosmological one: quasars participate in the expansion of the Universe. If this is the case then their large redshifts indicate that they are very far from us and therefore must be expending an enormous amount of energy to appear as bright as they do at these distances. For example consider the case of a quasar redshift of z=2, a value of redshift for some quasars that will be discussed later in this paper. Using the relativistic formula for redshift (the non-relativistic formula gives v>c for z>1) we obtain for recession velocity;



v=2.4x105 km/sec.

To get an approximate distance we use the currently accepted value of the Hubble parameter of 71 (km/sec)/Mpc, which yields a distance of 3400 Mpc or 11 billion light years. The fact that these objects can be observed at this kind of distance indicates that a typical quasar produces about 1000 times as much energy as an average spiral galaxy.

Not only do quasars emit energy at enormous rates but this energy comes from objects that are no larger that a few light years across. This defines the energy problem for quasars; how to generate 1000 times the energy of a galaxy in a region only a few light years across. This energy problem was one of the motivations for seeking a non-cosmological explanation. However it has been found that gravitationally fueled accretion of matter into massive black holes in the dense cores of galaxies can account for the necessary power (Silk 1989 p.266-269).

This interpretation of quasar redshifts is consistent with the Big Bang Theory and current theories of galaxy evolution. Quasars are all very distant objects with correspondingly large look-back times. Quasars are thought to represent early stages of galaxy formation and that the origin of their redshifts is similar to the origin of redshifts in distance galaxies. The redshift of their spectra is a cosmological Doppler shift, which results from their recession at high velocities over great distances. Gravitational redshift of light as it emerges from deep potential wells such as those created by massive black holes has been proposed as the origin of the large redshifts observed. However this hypothesis has been invalidated based on evidence of spectral emission lines being narrower than would be expected in such strong gravitational fields (Silk 1989, p.256-263). Also Compton scattering (inelastic scattering of photons off of free electrons) was rejected as the origin of the redshifts because there was no reasonable set of conditions where this effect could account for such high redshifts (Zhang 2008). So today the vast majority of astronomers take the view that quasars are indeed at great distances and their redshifts arise from the expansion of the Universe between the times of their creation early in the history of the Universe to the present.

However evidence began to accumulate that contradicted the conventional view that quasars are extremely distant. In 1966 Arp noticed that some high redshift quasars fell close to, and aligned across, some particularly disturbed galaxies at much lower redshifts. If they were really associated with relatively nearby galaxies they would themselves have to be relatively nearby (Arp 1987, p.7). It was concluded that these quasars had been ejected in some kind of explosive process. The large redshifts associated with these quasars was considered be intrinsic to the quasar and not related to velocity. In fact it was proposed that the redshift evolved with time as quasars were ejected, with high redshift occurring right after ejection gradually changing to lower redshift with time as the quasar moved away from the parent (Arp 1981).

In addition there are some cases of objects with widely differing redshifts but showing the appearance of direct interaction. Examples include both high redshift quasars connect to low redshift galaxies by luminous filaments or arms and pairs of discordant redshift galaxies similarly connected. Some well known cases include Stephans’ Quintet, MGC 4319 with the quasar Markarian 205 and NGC 7603. The reality of the apparent associations has not been absolutely established. The resolution of this question awaits further statistical analysis.

So here we come to the crux of the controversy. The reality of apparent associations of high redshift quasars with low redshift galaxies and the periodicity of redshifts depends upon the statistical methods used to analyze the data. There is statistical evidence favoring the non-cosmological nature of quasars and galaxies with large redshifts and other statistical arguments showing no significant association of quasars with low z galaxies and no significant periodicity of redshifts.

In this paper I will discuss evidence both for and against the non-cosmological interpretation of quasars and galaxy redshifts including analysis of the database provided by SDSS. I will discuss the ejection theory of quasars as an explanation for the apparent association of quasars with galaxies and the apparent periodicity or preferred values of redshift. I will also touch on possible physical mechanisms for intrinsic redshift (in a nutshell, there are none).

II. Non-Cosmological Redshift: Evidence and Models

Some years ago the hypothesis was advanced that some quasars were physically associated with companion galaxies. This hypothesis was initially based on analysis of the distribution of radio sources, including quasars, around active peculiar galaxies done by Halton Arp (1967). He found a statistically significant tendency of radio sources to be grouped around these active galaxies and that the radio sources appear to have been ejected from the active parent galaxies. In the case of quasars, Arp concluded that their redshifts must not be caused by velocity, but are intrinsic to the quasar. He based this conclusion on the absence of blue shifts or even relative blue shifting that would be expected if the shifts were largely Doppler velocity related shifts. The absence of any blue shifting indicated that ejection velocity was low enough not to contribute any significant component to the total redshift. Other radio sources associated with active galaxies had similar characteristics as quasars except to a lesser degree, which suggested to Arp that they may represent a sort of evolutionary sequence. Initially high redshift quasars were ejected by a parent galaxy. The redshifts of the quasars then relaxed to lower values with time. It was proposed that as the redshifts relaxed the quasars evolved into more normally behaved galaxies perhaps with a radio stage as part of their evolution (Arp 1967).

In 1978 Arp sampled bright spirals with companion galaxies. He had discovered previously that companion galaxies to larger central galaxies tended to have more quasars associated with them, presumably because the companions were younger and more active so ejected more quasars (Arp 1987 p.20). In this study he identified quasar candidates by photography with the 48” Schmidt camera at Mt. Palomar in two colors to identify stellar appearing objects with ultraviolet excess, which is characteristic of quasars. Then he used the spectrograph on the 5 meter telescope to obtain individual spectra to confirm their identity as true quasars.

From 34 predefined candidate galaxies, he found 13 cases where the quasars fell so close to the galaxies that the chances were, in individual cases, only about one in 100 of being accidentally associated. To find 13 such cases out of a limited number of trials implied fantastic odd against being a chance occurrence. Arp calculated about 10-17 probability of a chance occurrence (Arp 1987).

It is appropriate at this point to briefly discuss Arp’s statistical analysis approach since his data and statistics form the main body of evidence for the non-cosmological redshift hypothesis. The critical factor in determining the reality of an apparent association of high z quasars with low z galaxies is the probability that the observed proximity is just a chance alignment. Arp’s method of determining if quasar/galaxy associations were real was to look for a quasar around a candidate galaxy. When a quasar was found the probability of accidentally finding a quasar of this brightness close to the galaxy was calculated. In order to calculate the probability an all sky background density of 10 quasars per square degree down to magnitude B=V= 20 was adopted. This number was scaled down by a factor of 4 for each magnitude limit brighter so that for V=19 magnitude a background density of 2.5 per square degree was used. For example to calculate the number of background quasars down to magnitude 20 that fall within a 60” radius of a galaxy we calculate that within this radius there is 0.001 square degrees. The average background density of quasars down to magnitude 20 is about 10 per square degree. Therefore we would expect to find 0.001x10=0.01 background quasars within 60”. Thus the chance of finding a background quasar our little 60” circle is 1 in 100 (Arp 1987, p.15).

Now going back to the case of 13 close quasars/galaxy associations in a sample of 34 galaxies mentioned above; assuming a probability of finding an individual quasar is 0.01 or less, which is a reasonable probability based on the average background density as calculated above, to get the probability that 13 associations out of a sample of 34 galaxies would be a chance alignment we use the binomial distribution function;


This is the probability of an accidental association.

Arp’s method was to identify quasars around galaxies at successively larger radial distances until the probability of an individual quasar originating from the background distribution is greater than 0.01 using the background density of quasars scaled as discussed above. The aggregate probability of accidental association was then calculated using the binomial distribution as shown above. Other statistical analyses have also shown an excess of high redshift sources near low redshift galaxies (Lopez-Corredoira 2009).

Plenty of individual cases showing an apparent excess of quasars with high redshift near the center of nearby, low redshift, galaxies, mostly with AGN’s, have been discovered. In some cases the quasars are only a few arc seconds away for the center of the galaxies. Examples are NGC 613, NGC 1068, NGC 1097m NGC 3079, NGC 3842, NGC 6212, NGC 7541, NGC 7319 (galaxy quasar separation; 0.8”),, 2237+0305 (separation: 0.3”) and 3C 343.1 (separation: 0.25”) (Lopez-Corredoira 2009).

Cases of apparent association of objects with discordant redshifts have subsequently been discovered by Arp with probabilities of being chance alignments in the range of 10-8 to 10-19. If Arp is correct in his classification of the association objects as being quasars and his statistics are correct then it must be accepted that quasars are relatively low luminosity, nearby objects whose high positive redshifts are currently unexplained. The only way to avoid this conclusion is to contend that the average background density of distant quasars is somewhere around 20 times what it is currently measured to be (Arp 1981). We will see later in this paper that there is much skepticism regarding Arp’s results.

In 1979 Arp (Arp 1987 p.9-10) discovered that there were three quasars projected near the edge of the spiral galaxy NGC 1073. This was the first example of multiple quasars very close to a galaxy. He calculated the probability that three quasars would be observed by chance as close to the galaxy as 2x10-5. A further example is NGC 622 where a pair of quasars turned up very close to this galaxy. Especially significant is a filament of material that connects the quasar to the galaxy. This connecting filament suggests ejection from the galaxy as the explanation for the origin of this association. The discovery of alignment of quasars across disturbed galaxies and the similarity to radio sources whose alignment is caused by ejection from active galactic nuclei, suggests an ejection origin for quasars as well. The implication is that quasars are ejected from the nucleus of the associated galaxy.

Table 1 of Arp’s 1981 paper gave a summary of high z quasars associated with low z galaxies. Note that the redshift z values of the galaxies are given in km/sec consistent with a Doppler shift interpretation but the quasar z values are given directly. To convert the galaxy z values to direct z values divide them be the speed of light (3x105 km/sec). All of the galaxy redshifts are much smaller than the quasar redshifts, once again suggesting a non-velocity origin of the quasar redshifts.

In 1983 Arp published results of further statistical analysis for association of quasars around galaxies that were companions to large spiral galaxies (Arp 1987, p.24-25). Based on a sample size of 15 quasars he found that they concentrate at radial distance between 7 and 20 kpc. According to his analysis the density of quasars at these radial distances from their associated galaxies exceed by more than 20 times the measured density of quasars away from such galaxies.

Another example of quasars clustering around galaxies is the claimed excess of quasars around the Sculpter galaxy cluster (Arp 1987: 74).

Arp points to the distribution of high redshift quasars in the sky as further evidence for non-velocity redshifts. Results for the distribution of quasars in large areas of the sky indicate that high redshift quasars are concentrated preferentially in the half of the sky in the direction of the Local Group with a much lower concentration in the other half of the sky toward the Virgo Supercluster. This shows that many of the known quasars come from galaxies that are relatively close to us (Arp 1987: 69-70)

In addition to statistical evidence of association of high z quasars to low z galaxies, apparent physical connections between quasars and galaxies of discordant redshifts have been observed. Photographs of the quasar PKS 1327-206 plainly show a luminous filament connecting the quasar to a galaxy despite z=0.018 for the galaxy and z=1.17 for the quasar. Arp proposes that, since this quasar is quite bright in apparent magnitude, this galaxy quasar pair could be close to us and the quasar expelled by the galaxy (Arp 1987, p.38). Other examples of quasars connected to companion galaxies by luminous filaments are NGC 5297/96m NGC 5682 and NGC 7413. In all of these cases the redshift of the quasar is greater than that of the galaxy to which it is apparently connected. As will be discussed later, most astronomers take a very skeptical view regarding the reality of these connections.

Another important example of a high redshift quasar (zc=21,000km/sec) connected to a low redshift galaxy (zc=1700 km/sec) is the case of NGC 4319 and the quasar Markarian 205. Image processing work done by Jack Sulentic on plates made using the Mt. Palomar 5 meter telescope clearly showed a luminous bridge connecting the quasar to the galaxy. This luminous bridge continues straight back towards the galaxy’s nucleus and was perhaps indicative that the quasar emerged from the nucleus of this highly disturbed galaxy (Arp 1987, p.33).

Data from Chandra and XMM-Newton show many discrete powerful x-ray emitting quasars with a wide range of redshifts close to the nuclei of spiral galaxies. The data again suggests these sources are physically associated with the galaxies and are in the process of being ejected from them. A specific example is the x-ray emitting quasar very close to the nucleus of NGC 7319. z=2.114 for the quasar and z=0.0225 for NGC 7319 (Galianni 2005).

Other examples of where filaments/bridges/arms apparently connect objects of widely differing redshifts are NGC 3067/3C232, NGC 3628, ESO 1327-2041 connect to quasar 1327-206, 4C17.09, UGC 892 and others (Lopez-Corredoira 2009).

There is evidence that, in addition to quasar/galaxy pairs, some interacting galaxy pairs also show discordant redshifts, but to a lesser extent than quasar/galaxy pairs. An example is the Seyfert galaxy NGC 7603 and its companion. It is generally accepted that the companion is physically connected to NGC 7603 but the redshift of the larger galaxy corresponds to a velocity of 8700 km/sec and the smaller companion has a redshift of 17,000 km/sec (Arp 1982).

Physical association of objects with discordant redshifts is not the only line of evidence claimed for non-cosmological redshifts. Other evidence, which is difficult to explain using the conventional interpretation, is the periodicity of redshifts. In a homogeneous and isotropic universe we expect the redshift distribution of extragalactic objects to approximate a continuous and aperiodic distribution. There should be no preferred values of redshift.

Early studies of quasar absorption line redshifts indicated a preferred value of redshift z=1.95. This redshift was considered to be non-cosmological in nature and intrinsic to the quasars. Absorption by the interstellar medium was ruled out as a cause in this case. This conclusion was based on a very limited sample (7 quasars) and no mechanism was proposed to preferentially produce this value of redshift (Burbidge 1967).

More detailed studies by Karlsson reported a periodicity for quasars of 0.089 of the function log(1-z) for emission line redshifts of quasars. This conclusion was based on analysis of 574 quasar redshifts. He claims that this periodicity is real and not a result of selection effects as earlier periodicity results were shown to be. When quasars were associated with low redshift galaxies then the z relative to the rest frame of the galaxy must be used. The Karlsson model predicts peaks lying at z values of 0.061, 0.30, 06.0, 0.96, 1.14, 1.96 and so on. The Karlsson formula predicts the peak observed by Burbidge (1967; Karlsson 1977; Lopez-Corredoira 2009).

Karlsson acknowledges the difficulty in explaining this periodicity using the conventional cosmological model for redshifts but notes that periodicity is not conclusive evidence against the cosmological interpretation of redshifts. The peaks might correspond to active periods in the history of the Universe. The non-cosmological interpretation of quasar redshifts provides no really convincing explanation for periodicity either (Karlsson 1977).

As mentioned a number of times earlier in this paper, the hypothesis for associations of quasars with active galaxies was that quasars were ejected from the nuclei of the galaxies. An evolutionary hypothesis proposed by Arp (1967) was that compact objects were ejected from galactic nuclei, possibly at very high velocities, with initially very high density, high temperature and large redshifts (quasars and BL Lac objects). As they age, stars appeared as the redshift decreased. As the redshift decreased the luminosity increased. Finally, extended halos or spiral features appeared, and the non-cosmological redshift disappeared. Most quasars then are not at cosmological distances and are more like bright supergiants in luminosity than the incredibly luminous beacons that can be observed over billions of light years distance (Keal 2003).

Narlikar and Arp proposed an alternative cosmological model to the Big Bang to account for the evidence of intrinsic redshifts along with the large body of evidence for cosmological redshifts. Their proposal goes back to Hoyle’s steady state universe by assuming a matter filled, flat and static universe (Narlikar 1993).

According to their hypothesis, all particle masses scale with space and time according to m=at2 where t is time and a is a constant. This relation derives from a flat space-time cosmology where light travels without spectral shift. The observed cosmological redshift is a consequence of the systematic increase in particle mass with time because the wavelength of light emission scales inversely with mass. Hoyle explains the cosmological redshift in terms of look-back time to a distant galaxy, which shows it at an earlier era when its particle masses were smaller and the wavelength of emitted light longer (Narlikar 1993).

Redshifts that arise from differences in age could solve the periodicity problem. Material emerging from the zero mass, quantum mechanical realm may do so in discreet bursts spaced at discreet intervals of time. This could lead to certain preferred values of redshift (Narlikar 1993).

The excess of quasars around active galaxies can be described as above by an ejection mechanism from active galactic nuclei. This would be followed by evolution from the high z quasar state to low z normal galaxies with the intrinsic redshift gradually disappearing and giving way to a conventional velocity based redshifts. It’s still not clear to me how this hypothetical cosmology explains intrinsic redshift. Perhaps invoking the increase in particle masses with time and the inverse relation between emitted wavelength and mass as described above would provide an explanation.

Accounting for the observed properties of the microwave background is a problem with Nalikar and Arp’s hypothesis. They envision the microwave background as originating from the occurrence of mini-bangs through time resulting in, for example, ejection of quasars from galactic nuclei and creation of the microwave background. Hoyle has shown that a mini-bang of ~5x1015 solar masses could simulate big bang nucleosynthesis. It is hard to imagine how discrete minibangs could lead to the prefect 3K blackbody background radiation that has been observed. Rather discreet minibangs could well lead to quantization of the microwave background energy distribution which has never been observed. Thermalization to the observed homogeneity and isotropy of the microwave background is a major problem with this approach (Narlikar 1993).

A proposed explanation for intrinsic redshift states that it is determined solely by the gravitational potential associated specifically with the objects in which the emitting sources are located. During the process in which quasars evolve into ordinary galaxies, fragmentation and formation of stars results in reduction of the gravitational potential wells through which emitted light must travel thus the redshifts become smaller (Ching-Chuan Su 2006). Silk (1989 p.259-261) pointed out the expected emission line broadening that should occur if gravitation redshifts were an important factor in the total redshift have not been observed.

I should mention here another non-cosmological hypothesis that seeks to account for redshifts. In principle, light could be redshifted by effects other than velocity related Doppler shift. One such effect could be the so called tired light effect. Quanta of light could lose energy by some mysterious mechanism as it traverses space from remote galaxies. This decrease in energy would result in a distance dependant redshift. There is no known physical cause for such a loss in energy but small insignificant effects on a local scale may be manifested over the vastness of interstellar space.

This tired light hypothesis is generally ruled however because it specifies a static universe. Any static universe model has difficulties in accounting for the cosmic microwave background and light element abundances because of the absence of a hot big bang creation event (Silk 1989 p.396-397).

III. Evidence Against Non-Cosmological Redshifts

The vast majority of the astronomical community supports the cosmological interpretation of redshift and that the cases of apparent association of objects with discordant redshifts are just random projections of background/foreground objects. This interpretation of redshift as being caused by recession velocity imparted by the expansion of the Universe and thus indicative of distance, is consistent with the tenants of the Big Bang Theory, which is the best explanation we have for the origin and observed structure of the Universe. Under this interpretation quasars are very distant and represent an early era in galactic evolution. The expansion of the Universe provides a natural mechanism for producing redshifts and a sensible explanation for the relation between distance and redshift.

An objection leveled at some of the past results regarding association of objects with discordant redshifts was that the statistics were based on small samples or of only a few galaxies selected from a large sample. In order to rigorously test the validity of these association using a very large sample size, Tang (2008) used the highest quality data from Sloan Digital Sky Survey (SDSS) to test the assertion that there are excess quasars around nearby active galaxies. The sample size was 190,591 galaxies and 15,747 quasars. Data were compared with the ejection models and it was found that the radial distribution of quasars from active galaxies was fully consistent with randomly distributed quasars and galaxies and incompatible with ejection models. There was no significant indication of excess numbers of quasars in the regions around active galaxies (Tang 2008).

Tang’s (2008) method of analysis consisted of statistical test were run using different assumption; randomly distributed quasars and galaxies, randomly ejected quasars in which the quasars were produced by ejection from randomly chosen galaxies with a uniformly distributed age from 0 to 108 years and three different ejection velocities (11,000, 40,000 and 80,000 km/sec). 200 computer simulations for each set of conditions were run and the results plotted as the expected number of associated pairs vs. projected distance between high z quasars and active galaxies. The actual data fit the randomly distributed quasar model the best. The conclusion is that there is no solid evidence in SDSS data supporting the non-cosmological quasar hypothesis.

Tang (2008) points out that there are many solid examples which can be explained only if quasars are at cosmological distances. Observed gravitational lensing of quasars produced by foreground galaxies, where quasars must be much more distant than the lensing galaxies, which themselves have redshifts of the order of z>=0.5, is one example. Lensing would be expected to create an apparent excess of quasars in the direction intervening galaxies (Keal 2003). Also the presence of the Lyman-alpha forest and Gunn-Peterson trough seen in high redshift quasars indicate that light from these quasars had to pass through high redshift interstellar medium with large integrated neutral hydrogen densities (Tang 2008).

There is an active debate on the validity of redshift periodicities. In 2002 Hawkins et al. used 2dF data to test the periodicity in log(1+Zquasar) and found no periodicity (Tang 2008). Arp et al. (2005) criticized Hawkins citing his failure to take into account the critical factors of distance and brightness of the parent galaxies. Arp pointed out that the myriad of very faint galaxies will produce many random alignments thus masking true periodicity. Tang (2005) used the large database of quasar and active galaxy redshifts from the SDSS database to test for periodicity, taking into account Arp's criticisms of Hawkins work. Arp suggested that bright quasars will be nearby and that by using redshift data for these quasars, periodicity relationships would be more apparent than by examining distant quasars and applying correction factors. Therefore a subsample of bright quasars with magnitudes less than 18.5 were analyzed. The results of this analysis showed no periodicity in log (1+z) or any other frequency and no connection between foreground active galaxies and high redshift quasars (Tang 2005).

Furthermore work by John Bachall et al. (Bachall 1992) showed absorption lines in the spectrum of Markarian 205 produced by Mg II ions in the intervening spiral galaxy NGC 4319. This result was consistent with the cosmological interpretation of the redshifts of both Markarian 205 and NGC 4319, according to which the two objects are a chance alignment that happen to be projected close to each other in the sky and the light from Markarian 205 passes through the disk and halo of NGC 4319 (Bachall 1992). In addition further observations of distant quasars showed that they lie within galaxies with the same redshift, providing strong evidence that quasars are something that happens in the nucleus of galaxies (Keel 2003

IV. Summary

The redshift controversy has almost disappeared from sight. The overwhelming evidence for the Big Bang Theory, which requires that redshifts be interpreted as caused by velocity of recession and indicative of distance in an expanding universe, has caused the vast majority of astronomers to discount alternative non-cosmological explanations of redshifts.

I agree that the conventional view that the redshifts are cosmological in nature and that the existing Big Bang Theory is a superior description to Narlikar and Arp’s cosmology based on a static universe with intrinsic non-velocity related redshifts. Their approach is just too contrived especially regarding the homogeneity, isotropy and blackbody nature of the microwave background radiation. Another weakness of the non-cosmological redshift hypothesis is the lack of any reasonable physical process to produce redshifts. However one cannot help to at least consider that, based on the extensive studies done by Arp and others, there may be some degree of intrinsic redshift present in the spectra of quasars. Perhaps some of the quasar/galaxy interactions are real. It may be that observed redshifts are a combination of cosmological and intrinsic components. Intrinsic redshifts may damp out so quickly that they are rarely observed.

Despite the decline of non-cosmological redshift models, this debate served a very useful purpose for the advancement of cosmology as a whole. It caused astronomers to closely examine their data and assumptions under the paradigm of the Big Bang Model in response to Arp’s revolutionary proposal of non-cosmological redshifts. In this particular case it seems that the existing Big Bang paradigm was vindicated chiefly because of statistical evidence and because the competing cosmological model that included intrinsic redshifts could not account for the observed properties of the Universe as naturally and simply as the Big Bang.

The Big Bang cosmology has problems of its own however. Astronomers criticized the intrinsic redshift proposal because it did not provide a convincing physical mechanism to produce redshifts. The Big Bang has dark matter and dark energy. Despite strenuous experimental efforts nobody has yet detected dark matter. A paradigm changing revolution may still be in the future for the Big Bang Theory.


Arp, H. 1967, ApJ, 148, 321-365
Arp, H. 1981, ApJ, 250, 31-42
Arp, H. 1982, ApJ, 263,70
Arp, H. 1987, Quasars, Redshifts and Controversies, (Berkeley, CA: Interstellar Media)
Bachall, J.N., Januzzi, B.T., Schneider, D.P., Hartig, G.F., Jenkins, E.B. 1995, ApJ, 398, 496-500
Burbidge, G.R. & Burbidge, E.M. 1967, ApJ, 148, L107
Ching-Chuan Su 2006, arXiv:physics/0608164v1 [physics.gen-ph]
Galianni, P., 2005, ApJ, 620:88-94
Keal, W. 2003, “Alternate Approaches and the Redshift Controversy,
Karlsson, K.G. 1977 A&A, 58, 237-240
Lopez-Corredoira, M. 2009, arXiv:0910:4297v1 []
Narlikar, J.& Arp H. 1993, ApJ, 405, 51-56
Silk, Joseph 1989, The Big Bang, (New York, W. H. Freeman and Company)
Tang, S.M., Zhang, S.N. 2005, ApJ, 633, 41-51
Tang, S.M., Zhang, S.N. 2008, arXiv:0807.2641v2 [astro-ph]
Zhang, P. 2008, arXiv:0802.2417v2 [astro-ph]