IN STARRY SKIES

Evening Post, Volume CIX, Issue 117, 21 May 1925, Page 16

IN STARRY SKIES

THE GLOBULAR CLUSTERS

PUZZLES ON A HUGE SCALE

(By " Omega Centauri.")

The distribution of the globular clusters in the sky is very remarkable, and probably extremely significant. Ninetyfive of these vast systems are known, and nearly all of them lie in a single hemisphere, of which the pole is in the direction of the great star clouds of Sagittarius. This is also the direction in which Novae are found most frequently. The large Spiral Nebulae, on the other hand, are most abundant in the opposite half of the sky. The globular clusters are not evenly spread over the hemisphere in which they occur. None at all are found within five degrees of the central plane of the Galaxy, although they are m^st abundant only a few degrees further from it. There is something very suggestive about this peculiar distribution, but- to understand it we must know something about the distances of these wonderful systems from us. Thejr are nil believed to be exceedingly large and very far away. Every one is far too distant for any direct determination of its parallax to be possible. A simple analogy may make this clearer. Suppose' that all our instruments are contained in a room with a single widow one yard wide, and we are required to determine the distance of a light on a distant mountain. Wo shall suppose that tbe distance is 117 miles, but that we know nothing of it. From the two sides of the window the light appears in slightly different directions, but the two are so nearly -parallel that the difference is difficult to detect, and still more to measure. The angle is, in fact, one second. Practically every determination of stellar distance is very much more difficult than this. In order to increase the angle between the .two lines a surveyor increases his base line. An astronomer cannot do so indefinitely. The longest base line available is the diameter of the earth's orbit, which is less than 186 million miles. Half of this is taken as the astronomical unit, and its exact 'value is considered to be not far from 92,860,000 miles. Even this unit is found to-be far too small "as soon as we leave the solar system, . 5, ™uf h larger one is derived from it. This is the distance from which an astronomical unit subtends one second of arc. This is called a parsec. It is over 19 .billion ■ miles. There. is no known star witbiir a parsec of the sun. Even Alpha Centauri is a third as far off again. Yet the most skilful modern observers, with the most perfect instruments of the present day, succeed in measuring distances 50 times as great as that of Alpha. Centauri. Over 1200 parallaxes have'thus been determined. To make the conditions ■. as difficult, in our terrestrial analogy' we should have to cut down the width of the 1 window to about an inch. .

Nearly .all the stars,- all the globular clusters, and all the Spiral Nebulae are far too distant for any such direct method to have the slightest chance of success. Fortunately in the case of the stars, and also in the case of the globular clusters, several different indirect methods have- been found. to give fairly concordant results. The measures arrived at, though acknowledged to be only approximate, are believed to be of the right order of magnitude. From their appearance the globular clusters would be judged to be at widely different distances. They are wonderfully alike in structure but differ very greatly in apparent size. When the faint ones are examined with a highpower they_ look just like the brighter ones did with a lower. It seems quite probable that, on the whole, those which appear smaller and fainter are further away than others which look larger and brighter.' Again, the apparent°:magmtudes of the stars in different clusters are related to the apparent diameters of the clusters in such a way as to suggest that the apparent differences are due chiefly to differences in their distances from us. If this is so, we have a ready means of arranging the clusters approximately, in order of relative distance Buf this gives no indication, of the absolute scale. To determine this we must settle the real distance of at least one cluster.

Now many o f the brighter stars in some clusters ' can have their spectral types determined. It seems probable that, even in these .distant systems, the stars or any particular type, will be of somewhat the same average absolute luminosity as those of the same type in the neighbourhood of the sun. The absolute magnitude of the latter is-known from their apparent brightness and calculated distance. ' Assuming ' that similar stars in the cluster are equally bright, we can, therefore, from their apparent magnitude, deduce the distance of the cluster. One difficulty arises : are the cluster stars giants or dwarfs? Probably most of those which we see are giants, but it is unnecessary to make this assumption. The difficulty can be avoided by choosing the hottest stars, which are known as type B, for amongst them there is little or no difference between giants and dwarfs. This method is found to confirm the. results arrived at m other ways. A .still more definite determination has been made by Shapley from a study of the Cephefd Variables. In some clusters these are very numerous. They change periodically in brightness by one. or one and a half ■magnitudes. Sonio go Ihrough their cycJe in a tew hours; none take moro than 50 days. They aro believed to bo pulsating stars. Misg Lcavitt made a special study of the Lesser Magellanic Cloud, m the course of which she dis-covered-lhal the periods of the variables in it bore a definite relation to their apparcnt _ brightness. Now ■ all the stars in tho_ Cloud may be considered to be approximately at the same distance from ■us, as the Cloud is more than a dozen times the length of its diameter from us. Iho relation found by Miss Leavitt is, therefore, one between absolute luminosity and period. Now some galactic Cepheids aro at known distance from us. Iheir absolute luminosity is determined from their apparent brightness. If the same law holds for all Cepheids, we have a clue to their distances, and, therefore, to the distances of the clusters which contain them.

In some clusters no Cepheids have been found. Their distances must be assigned by other methods. The brightest stars in a cluster which contains Cepheids are found by photography to be on an average about ten times as luminous as .1 typical first magnitude star. If the same is true in a cluster which does not contain variables its distance can _be assigned. The results are surprising. The two nearest clusters arb Omega Centrum and 47 Toucani, -with, distances of about 21,000 and 22.000 light, years respectively. I\]3, in Canes Venatici, near the North gal.-itic pole, is ove» 45,000 light-years away, and is so vast that light would lake 470 years to cross from one side of it to the" other. Even with this vast extension, the stars in is arc very much more crowded than thost round the sun. Professor Dingle estimates (hat in the central portion there are at least 15,000 stars in a space equal to that round our t».in, which cuntainii but four or. five ttnvs. The motli diitaii!; tduster jqi. ssludjs'd is ft.G.U.

700, and which Shapley considers to be 220,000 light years away. Some of these systems contain hundreds of thousands of stars.- In addition to the motions of the individual stars with regard to one another, most clusters have a rapid motion as a whole. Stipher has measured the radial velocities of 15 globular clusters, and finds the surprising range of from zero to 256 miles a second. The average for the whole 15 is 78 miles a. second. Strange to say, . only four out of the fifteen are receding from us. Taking this into consideration at the same time as the other puzzling fact that nearly all the spiral nebulae are receding from us, it appears possible that our whole galactic system is approaching the great system of the globular clusters, and moving away from that of the spiral nebulae. If so, it must be moving in the direction of the lure from the Solar System to the star, clouds of Sagittarius..

The globular clusters are magnificent objects when seen through a telescope, but the excellent photographs now published make it possible to study them even more successfully, by one's own fireside. One of the strangest peculiarities seen in many of the photographs is the apparent streaming of the outlying stars in graceful lines and curves. Owing to their high southern declina* tions, neither Omega Centauri nor 47 Toucani has been photographed with a large reflector. Excellent photographs have, however, been taken with the 24in Victoria refractor at. the Cape Observatory, and also by Mr. Franklin Adams.