In 1750 Thomas Wright of Durham, England, published his cosmological credo in a handsomely designed volume augustly entitled An Original Theory or New Hypothesis of the Universe Founded Upon the Laws of Nature, and Solving by Mathematical Principles the General Phaenomena of the Visible Creation; and Particularly the Via Lactea. His ideas would have likely been buried in obscurity if not for the fact that others widely disseminated what they thought he meant, which led to his reputation as the first to postulate a model of the Via Lactea-Milky Way as a disk of stars.

Once the solar system was found to be organized according to Newton’s laws, a number of investigators began to consider whether the stars as well were arranged in a particular way. And, if so, what was the Sun’s position within such a structure? Pondering cosmic models was a favorite pastime for Wright as a youth. The third son of a well-to-do carpenter, he taught himself mathematics and was so obsessive about studying astronomy that his father at one point burned his books. Though limited in formal education, he was always concocting grand intellectual schemes and wrote that his dream was to produce “an integrated picture of natural and supernatural, of creation and Creator.” Apprenticed to a clockmaker at the age of thirteen, Wright later taught navigation to seamen. His abilities were noted by the aristocracy, and he eventually made a comfortable living giving lectures and private instruction to noble English families. It was through such aristocratic benevolence that he was largely able to publish his most famous work, lavishly illustrated with thirty-two engravings, at the age of thirty-nine.
The central theme of his Original Theory, composed as a series of letters, was that the Sun was just one of many stars revolving around a common center of gravity. This was a rational assumption, as Edmond Halley had recently detected motion in the so-called fixed stars, and a rotation would prevent the stars from gravitationally collapsing toward one another. At one point in his text Wright illustrates the stars moving in a ring (or series of rings), much like the rings of Saturn. But, strongly guided by religious views, he preferred to think of the Milky Way as a spherical shell of stars, with the Sun off to one side and the Eye of Providence, the “agent of creation,” residing in the center. Wright’s diagram of the Milky Way as a flat layer of stars, a familiar illustration in astronomy textbooks, was actually a first step in helping his readers picture this shell - so vast in size that the small segment in which we reside would appear to be a flat plane. “I don’t mean to affirm that [the plane] really is so in fact,” he writes, “but only state the question thus to help your imagination to conceive more aptly what I would explain.”
Despite his awkward mix of theology and science, Wright does deserve credit for his insight that the Milky Way was an optical effect, the result of the solar system being immersed in an assembly of stars. He introduced the idea, now viewed as common sense, that our position in space affects how we perceive our celestial environment. The Milky Way appears as a band, he mused, because we observe a thin layer of stars edge-on, its combined light producing the milk-white appearance. Such a structure also explains why, when looking perpendicular to the band, stargazers see fewer stars.
Wright went on to speculate that the cloudy spots then being observed by astronomers in greater numbers might be external creations, “bordering upon the known one, too remote for even our telescopes to reach.” This idea was amplified in 1755 by the philosopher Immanuel Kant and independently suggested six years later by the German mathematician Johann Heinrich Lambert in his Cosmological Letters.

from

An Original Theory or New Hypothesis of the Universe (1750)

By Thomas Wright

Letter the Seventh
. . When we reflect upon the various aspects and perpetual changes of the planets, both with regard to their heliocentric and geocentric motion, we may readily imagine that nothing but a like eccentric position of the stars could any way produce such an apparently promiscuous difference in such otherwise regular bodies. And that in like manner, as the planets would, if viewed from the Sun, there may be one place in the universe to which their order and primary motions must appear most regular and most beautiful. Such a point, I may presume, is not unnatural to be supposed, although hitherto we have not been able to produce any absolute proof of it.
This is the great Order of Nature, which I shall now endeavor to prove and thereby solve the phenomena of the Via Lactea [Milky Way]; and in order thereto, I want nothing to be granted but what may easily be allowed, namely that the Milky Way is formed of an infinite number of small stars.
Let us imagine a vast infinite gulf, or medium, every way extended like a plane and enclosed between two surfaces nearly even on both sides, but of such a depth or thickness as to occupy a space equal to the double radius or diameter of the visible creation, that is to take in one of the smallest stars each way, from the middle station, perpendicular to the plane’s direction, and as near as possible according to our idea of their true distance.
But to bring this image a little lower, and as near as possible level to every capacity, I mean such as cannot conceive this kind of continued zodiac, let us suppose the whole frame of nature in the form of an artificial horizon of a globe. I don’t mean to affirm that it really is so in fact, but only state the question thus to help your imagination to conceive more aptly what I would explain. [Figure 21.1] will then represent a just section of it. Now in this space let us imagine all the stars scattered promiscuously, but at such a distance from one another as to fill up the whole medium with a kind of regular irregularity of objects. And next let us consider what the consequence would be to an eye situated near the center point, or anywhere about the middle plane, as at the point A. Is it not, think you, very evident that the stars would there appear promiscuously dispersed on each side, and more and more inclining to disorder as the observer would advance his station towards either surface and nearer to B or C, but in the direction of the general plane towards H or D, by the continual approximation of the visual rays crowding together as at H between the limits D and G, they must infallibly terminate in the utmost confusion. If your optics fails you before you arrive at these external regions,

Figure 21.1

only imagine how infinitely greater the number of stars would be in those remote pans, arising thus from their continual crowding behind one another, as all other objects do towards the horizon point of their perspective, which ends but with infinity: Thus, all their rays at last so near uniting must meeting in the eye appear, as almost, in contact and form a perfect zone of light; this I take to be the real case and the true nature of our Milky Way, and all the irregularity we observe in it at the Earth, I judge to be entirely owing to our Sun’s position in this great firmament and may easily be solved by his eccentricity and the diversity of motion that may naturally be conceived amongst the stars themselves, which may here and there, in different parts of the heavens, occasion a cloudy knot of stars, as perhaps at E.

Figure 21.2

But now to apply this hypothesis to our present purpose and reconcile it to our ideas of a circular creation and the known laws of orbicular motion, so as to make the beauty and harmony of the whole consistent with the visible order of its parts, our reason must now have recourse to the analogy of things. It being once agreed that the stars are in motion, which, as I have endeavored in my last letter to show is not far from an undeniable truth, we must next consider in what manner they move. First then, to suppose them to move in right lines, you know is contrary to all the laws and principles we at present know of; ... it must of course be the other, i.e. in an orbit; and consequently, were we able to view them from their middle portion, as from the Eye seated in the center of [Figure 21.2] we might expect to find them separately moving in all manner of directions round a general center, such as is there represented. It only now remains to show how a number of stars, so disposed in a circular manner round any given center, may solve the phenomena before us. . . .The first is in the manner I have above described, i.e. all moving the same way, and not much deviating from the same plane, as the planets in their heliocentric motion do round the solar body. . . .
The second method of solving this phenomena is by a spherical order of the stars, all moving with different direction round one common center, as the planets and comets together round the Sun, but in a kind of shell or concave orb. . . .

Hence we may imagine some creations of stars may move in the direction of perfect spheres, all variously inclined, direct and retrograde; others again, as the primary planets do, in a general zone or zodiac, or more properly in the manner of Saturn’s rings, nay. perhaps ring within ring, to a third or fourth order, as shown in [Figure 21.3], nothing being more evident than that if all the stars we see moved in one vast ring, like those of Saturn, round any central body or point, the general phenomena of our stars

Figure 21.3

would be solved by it. . . . Not only the phenomena of the Milky Way may be thus accounted for, but also all the cloudy spots and irregular distribution of them; and I cannot help being of [the] opinion that could we view Saturn through a telescope capable of it, we should find his rings no other than an infinite number of lesser planets, inferior to those we call his satellites: What inclines me to believe it is this; this ring or collection of small bodies appears to be sometimes very eccentric, that is more distant from Saturn’s body on one side than on the other and as visibly leaving a larger space between the body and the ring; which would hardly be the case if the ring, or rings, were connected or solid, since we have good reason to suppose it would be equally attracted on all sides by the body of Saturn, and by that means preserve everywhere an equal distance from him; but if they are really little planets, it is clearly demonstrable from our own in like cases, that there may be frequently more of them on one side than on the other, and but very rarely, if ever, an equal distribution of them all round the Saturnian globe.

How much a confirmation of this is to be wished, your own curiosity may make you judge, and here I leave it for the opticians to determine. I shall content myself with observing that Nature never leaves us without a sufficient guide to conduct us through all the necessary paths of knowledge; and it is far from absurd to suppose Providence may have everywhere throughout the whole universe, interspersed modules of every creation, as our Divines tell us, Man is the image of God himself.

Thus, Sir, you have had my full opinion, without the least reserve concerning the visible creation, considered as part of the finite universe; how far I have succeeded in my designed solution of the Via Lactea, upon which the theory of the whole is formed, is a thing that will hardly be known in the present century, as in all probability it may require some ages of observation to discover the truth of it. . . .

The Discovery of Other Galaxies

The universe as we know it was revealed to astronomers on New Year’s Day 1925. The man responsible, Edwin Hubble, was not present. His historic paper, primly entitled “Cepheids in Spiral Nebulae,” was read to the thirty-third annual meeting of the American Astronomical Society in Washington, D.C., by Henry Norris Russell of Princeton. Reluctant to divulge his findings too soon, Hubble had to be persuaded to release a preliminary report at the conference. The announcement changed the face of the universe—the Milky Way was suddenly humbled, becoming just one of a multitude of galaxies residing in the vast gulfs of space. Hints of the Milky Way’s true place in the universe had been cropping up for years, but Hubble’s observations provided the evidence that at last convinced the community at large.

Establishing that certain nebulae were sister galaxies of the Milky Way had a long and tumultuous history. In the eighteenth century the great British astronomer William Herschel discovered hundreds of nebulae scattered over the celestial sphere and, like Immanuel Kant earlier, first thought of them as island universes, separate congregations of stars. But over time support for this view waned. The work of William Huggins and other spectroscopists in the nineteenth century showed that many nebulae were gaseous in nature, leading to the popular view that all nebulae were members of the Milky Way, possibly other solar systems or new star clusters in the making. “A good many converging lines of evidence,” wrote historian Arthur Berry in 1898, “indicate the probability that [the celestial] bodies should be regarded as belonging to a single system.” In 1885, for example, there was a dramatic flaring in the Andromeda nebula. This nova, at the height of its brilliance, rose to the seventh magnitude. If Andromeda were an external universe, far beyond the Milky Way’s borders, that nova had to be shining with the energy of some fifty million suns. Unaware as yet that a star could explode as a supernova, astronomers considered such an output preposterous. Nineteenth-century historian Agnes Clerke remarked that it would “have been on a scale of magnitude such as the imagination recoils from contemplating.” Ten years later, another bright nova was spotted in NGC 5253, a spiral nebula in the constellation Centaurus [1]. To Clerke, the idea that nebulae were other universes had “ceased to exist.”

Despite these declarations, a few lone astronomers began gathering evidence that Andromeda and other spiral nebulae were remote galaxies of stars after all. In 1917 at Mount Wilson Observatory George Ritchey, closely examining a photograph, spotted a previously unseen star in NGC 6946, but one far dimmer than the 1885 flare-up in Andromeda. Heber Curtis at California’s Lick Observatory found other cases of faint novae in spiral nebulae, one in NGC 4257 and two others in NGC 4321. This suggested there were two types of flaring: the bright novae of 1885 and 1895 were exceptions rather than the rule and reopened the possibility that the spiral nebulae were indeed islands of stars. Curtis had been conducting a photographic survey of nebulae and was already convinced. In 1914, in an in-house Lick report, he remarked that the spirals are “inconceivably distant, galaxies of stars or separate stellar universes so remote that an entire galaxy becomes but an unresolved haze of light.” He even found the reason that spiral nebulae tended to crowd around the poles of the Milky Way and were not seen in its plane. He photographed some spiral nebulae edge-on and noticed dark lanes of obscuring matter. Such dusty material in the disk of our own galaxy would of course, reasoned Curtis, hide our view of the distant universe along that direction.
In 1920, under the sponsorship of the National Academy of Sciences in Washington, D.C., Curtis squared off with Harvard’s Harlow Shapley, chief proponent of the view that the Milky Way was the universe’s sole galaxy, in what has come to be known in astronomy as the “Great Debate.” Shapley had one, very compelling rebuttal to Curtis’s arguments. At Mount Wilson, the Dutch astronomer Adriaan van Maanen was examining photographs of spiral nebulae taken at different times and was claiming to see the nebulae rotate, an effect impossible to see unless the nebulae were small, and hence fairly close, within the Milky Way itself. The debate was a draw.

In 1922 Ernst Öpik at the Dorpat Observatory in Estonia determined that the Andromeda nebula was some 1.5 million light-years distant, by assuming that its mass and luminosity were comparable to those of the Milky Way. But his paper received little notice. It was Hubble who provided the conclusive evidence when he directly determined the distance to Cepheid variable stars within Andromeda (M31) and the M33 spiral in Triangulum. It was the type of distance measurement in which astronomers had the most confidence.

A Rhodes scholar trained in law, Hubble had returned to graduate school and his favorite college subject, astronomy, in 1914 at the age of twenty-four and five years later became a member of the Mount Wilson Observatory staff, which gave him access to the 100-inch Hooker reflector. He was part of a select group in California that for several decades dominated astronomy’s discoveries in the far universe because of its employment of the world’s largest and best telescopes. In the fall of 1923, Hubble began a study of Andromeda, spotting two ordinary novae and a third faint star. Perusing the library of plates on Andromeda, going back to 1909, he came to realize that the third stellar object was a variable star. With more than sixty plates on hand, he plotted the star regularly rising and falling in intensity. Six nights at the telescope in February 1924 confirmed it: the star’s brightness rose rapidly that week, just like a Cepheid. Its period was thirty-one days, which, according to the period-luminosity relation established by Henrietta Leavitt and calibrated by Shapley, meant it was an extremely luminous star. But Hubble determined that this Cepheid, being only of eighteenth magnitude, had to be almost a million light-years away (285,000 parsecs) to appear so faint [2]. His estimate of Andromeda’s distance fell short of today’s value of 2 million light-years, but it still put the nebula far beyond the confines of the Milky Way [3]. Hubble eventually found other Cepheid variables, in both M31 and M33, that clinched his conclusion. He wrote Shapley, who recognized the overwhelming evidence and quickly conceded that he had been wrong about Andromeda. But van Maanen (whose rotation data were later found to be erroneous) resisted, which kept Hubble from making an immediate announcement. Shapley and others, though, convinced Hubble to have a paper summarizing his findings read at the 1925 astronomy meeting. There is no direct mention of external galaxies or a cosmological shake-up in his report; as was his style, Hubble kept his announcement low-key. Only in the American Astronomical Society’s report of the meeting was it noted that Hubble had brought “confirmation to the so-called island universe theory.”

Hubble devoted the rest of his professional life to the realm of the nebulae. He soon introduced a classification scheme to distinguish the various types. There are the E or elliptical galaxies, the smooth spheroidal bulges that run from round to oval in shape; the S or spiral galaxies, which consist of a central bulge in a range of sizes surrounded by a spiraling disk that can be tightly coiled or spread out wide; and lastly the irregular galaxies, such as the chaotic Magellanic clouds.

“Cepheids in Spiral Nebulae.”

Publications of the American Astronomical Society, Volume 5 (1935)

by Edwin P. Hubble

Messier 31 and 33, the only spirals that can be seen with the naked eye, have recently been made the subject of detailed investigations with the 100-inch and 60-inch reflectors of the Mount Wilson Observatory. Novae are a common phenomenon in M31, and Duncan [4] has reported three variables within the area covered by M33. With these exceptions there seems to have been no definite evidence of actual stars involved in spirals. Under good observing conditions, however, the outer regions of both spirals are resolved into dense swarms of images in no way differing from those of ordinary stars. A survey of the plates made with the blink-comparator has revealed many variables among the stars, a large proportion of which show the characteristic light-curve of the Cepheids [5].

Up to the present time some 47 variables, including Duncan’s three, and one true nova have been found in M33. For M31, the numbers are 36 variables and 46 novae, including the 22 novae previously discovered by Mount Wilson observers. Periods and photographic magnitudes have been determined for 22 Cepheids in M33 and 12 in M31. Others of the variables are probably Cepheids, judging from their sharp rise and slow decline, but some are definitely not of this type. One in particular, Duncan’s No. 2 in M33, has been brightening fairly steadily with only minor fluctuations since about 1906. It has now reached the 15th magnitude and has a spectrum of the bright line B type.

For the determinations of periods and normal curves of the Cepheids, 65 plates are available for M33, and 130 for M31. The latter object is too large for the area of good definition on one plate, so attention has been concentrated on three regions: around BD +41° 151, BD +40° 145, and a region some 45´ along the major axis south preceding the nucleus.

Photographic magnitudes have been determined from twelve comparisons with selected areas No. 21 and 45, made with the 100-inch using exposures from 30 to 40 minutes. This procedure seemed preferable to the much longer exposures required for direct polar comparisons with the 60-inch. It involves, however, a considerable extrapolation based on scales determined from the faintest magnitudes available for the selected areas.
Tables [51.1] and [51.2] give the data for the Cepheids in M33 and M31 respectively. No magnitudes fainter than 19.5 are recorded, because of the uncertainty involved in their precise determinations. The now familiar period-luminosity relation is conspicuously present.

Table 51.1

For more detailed investigation of the relation, the magnitudes at maxima have been plotted against the logarithm of the period in days. This procedure is necessary, not only because of the uncertainties in the fainter magnitudes, but also because most of the fainter variables at minimum are below the limiting magnitude of the plates. It assumes that there is no relation between period and range, for otherwise a systematic error in the slope of the period-luminosity curve is introduced. Among the brighter Cepheids of M33 the assumption appears to be allowable, for the ranges show a very small dispersion about the mean value of 0.8 magnitude. The average range and the dispersion are somewhat larger in M31, but the data are too limited for a complete investigation.

The curve for M33 appears to be very definite. The average deviation is about 0.1 magnitude, although a considerable systematic error is allowable in the slope. For M31 the slope is very closely the same but the dispersion is much greater, averaging about 0.2 magnitude. This is probably greater than the accidental errors of measurement.

Shapley’s period-luminosity curve for Cepheids, as given in his study of globular clusters, is constructed on a basis of visual magnitudes. It can be reduced to photographic magnitudes by means of his relation between period and color-index, given in the same paper, and the result represents his original data. The slope is of the order of that for the spirals, but is not precisely the same. In comparing the two, greater weight must be given the brighter portion of the curve for the spirals, because of the greater reliability of the magnitude determinations. When this is done, the resulting values of M - m are 21.8 and 21.9 for M31 and M33, respectively. These must be corrected by half the average ranges of the Cepheids in the two spirals, and the final values are then on the order of 22.3 for both nebulae. The corresponding distance is about 285,000 parsecs. The greatest uncertainty is probably in the zero point of Shapley’s curve.

Table 51.2

The results rest on three major assumptions: (1) The variables are actually connected with the spirals. (2) There is no serious amount of absorption due to amorphous nebulosity in the spirals. (3) The nature of Cepheid variation is uniform throughout the observable portion of the universe. As for the first, besides the weighty arguments based on analogy and probability, it may be mentioned that no Cepheids have been found on the several plates of the neighboring selected areas No. 21 and 45, on a special series of plates centered on BD +35° 207, just midway between the two spirals, nor in ten other fields well distributed in galactic latitude, for which six or more long exposures are available. The second assumption is very strongly supported by the small dispersion in the period-luminosity curve for M33. In M31, in spite of the somewhat larger dispersion, there is no evidence of an absorption effect to be measured in magnitudes.

These two spirals are not unique. Variables have also been found in M81, M101 and NGC 2403, although as yet sufficient plates have not been accumulated to determine the nature of their variation.

---

Footnotes

[1] NGC stands for New General Catalogue, published by J. L. E. Dreyer in by request of the Royal Astronomical Society. It extended the general catalogue published by John Herschel in 1864 and remains a standard for referencing deep-sky objects.
[2] 1 parsec = 3.26 light-vears.
[3] Because Hubble was not aware that there were two classes of Cepheids, with different period-luminosity relationships, his distance was off by a factor of about two. The error would be discovered in the 1950s. Öpik’s 1922 estimated distance to M31 turned out to be more accurate in the end.
[4] John C. Duncan, director of the Wellesley College observatory, who first spotted a variable star in M31 in 1922.
[5] A blink comparator allows a viewer to quickly alternate between two photographic plates taken of the same field at different times. The blinking proceeds so rapidly that a changing or moving object will stand out, while those that remain fixed appear still.