144 ANNUAL REPORT SMITHSONIAN INSTITUTION, 19 3 8 



investigations during the next 5 years increased the number to 40, 

 and a dozen or more have since been added to the list. The light- 

 curves are typical and the periods, with one outstanding exception 

 (P=175 days), range from 48 to 10 days. The data exhibit the fa- 

 miliar relation between periods and mean, or maximum, luminosities — 

 the famous period-luminosity relation — by virtue of which the 

 Cepheids furnish the best available criterion of great distances. 

 Wherever we find Cepheids, their periods indicate their intrinsic 

 luminosities (candlepowers) and, consequently, their apparent faint- 

 ness measures their distances. For instance, the known Cepheids in 

 M31, as indicated by their periods, average about 1,800 times as 

 bright as the sun, but they appear about 200,000 times fainter than 

 the faintest stars seen with the naked eye. 



The distance of M31 would be rather accurately determined by 

 this method if it were not for the uncertain effects of obscuration, 

 both in the galactic system and in the spiral. The main body of the 

 galactic system appears to be embedded in a widespread, tenuous 

 medium (often called the uniform layer), and, of course, many clouds 

 of dust and gas are scattered through the system, concentrating 

 toward the fundamental plane of the Milky Way. From our posi- 

 tion within the system, and close to the galactic plane, we observe 

 distant nebulae through the surrounding haze and between the 

 obscuring clouds. The absorption is greatest in and near the Milky 

 Way, and diminishes toward the galactic poles. But the effects are 

 not strictly uniform as we look around the sky at a given galactic 

 latitude, and the calculated corrections, at present, represent average 

 values which may not be precise for a particular direction. 



Now M31 is seen near the edge of the Milky Way at latitude 20°. 

 The average obscuration at this latitude is about 0.48 magnitude (55 

 percent) but the proper correction might be considerably different. 

 Actually, the large numbers of very faint, distant nebulae found in the 

 vicinity of M31 suggest that the region is more transparent than nor- 

 mal. A cloud, or real aggregation of the background nebulae, would 

 produce the same phenomenon, and it is quite possible that both effects 

 are present. However, an examination of the data leads to the pro- 

 visional conclusion that abnormal transparency is the dominating 

 factor, and, consequently, the effect will be partially compensated by 

 minor obscuration within the spiral. These remarks sufficiently 

 indicate that the precise effect of obscuration is an unsolved problem. 

 For the present, it is advisable to apply the average correction for 

 galactic latitude, and to recognize an uncertainty of the order of 10 

 to 25 percent in the resulting distance. 



The direct procedure is then to compare the many Cepheids in M31 

 with those in the Small Magellanic Cloud which serves as a unit of 

 extragalactic distance. Taking account of the average galactic ob- 



