196 ANNUAL REPORT SMITHSONIAN INSTITUTION, 192 8 



spaced, there are concentrations of stars here and there which, seeiv 

 from another part of the galaxj^ produce such beautiful effects as 

 the " star clouds " in Sagittarius or the " globular clusters " in Her- 

 cules and other parts of the sky. Within the galaxy, too, filling 

 great regions of space around and between some of the stars, are the 

 gaseous nebula3 both dark and bright. 



Keturning now to the extragalactic nebulae — the great gaseous ob- 

 jects and star clusters like islands in a three-dimensional ocean of 

 space beyond the Milky Way — we are indebted to Dr. E. P. Hubble 

 of Mount Wilson Observatory for much new knowledge concerning 

 them. In a recently published paper - he has given the results of a 

 careful study of 400 such nebulae. Some of the extragalactic nebulae 

 show no regularity of shape or structure. These form a subcjass by 

 themselves, and to this class belong the Magellanic Clouds, great 

 irregular star clouds visible from the Southern Hemisphere like de- 

 tached portions of the Milky Way, though actually as far away 

 again. Far more numerous than the irregular nebulae are those 

 liaving a definite shape or structure, the ellipsoidal and the spiral 

 nebulae. The spectra of the former are so similar to the solar spec- 

 trum that there is no room for doubt that the}^ are clusters of stars, 

 even though the individual stars can not be photographed. Possibly 

 the stars are being gradually condensed out of the gases of which 

 the nebulae were composed, and the residual gases act as envelopes 

 rendering the star images indistinct. 



Some of the nebulae are apparently at a transition stage between 

 ellipsoidal and spiral, while yet others display well-developed spiral 

 ai-ms. Tlie evidence seems strongly to point tow^ard an evolutionary 

 ])rocess, as a glance at the illustrations will make clear — the gradual 

 unwinding, the appearance of stars and star streams, the whole vast 

 process of the development of island galaxies. 



With this idea of progressive development in mind, the spiral 

 nebulffi are said to be of early, intermediate or late type, according 

 as they present the appearance of the uncondensed nebula in Plate 

 1, or intermediate forms between that and the well-developed, far- 

 fiung, stellar arms so clearly shown in Plates 4 and 5. 



The distances of some of the spirals have been determined in a 

 very interesting way. These spirals contain stars, known as Ce- 

 pheid variables, whose light is not steady but fluctuates with perfect 

 regularity, falling slowly from maximum to minimum and then ris- 

 ing rapidly to maximum. The liglit cycles usually have periods of 

 a fcAV hours or a few days. When studying similar stars whose 

 true brightness was known, Miss Leavitt of Harvard Observatory 

 discovered the fact that tlie longer tlie period of light variation the 



* Extragalactic Nebula. E. P. Hubble. Astrophysical Journal, December, 1926. 



