124 ANNUAL. REPORT SMITHSONIAN INSTITUTION, 19 3 2 



omer, Holwarda, who observed that it appeared and disappeared 

 within a period of about 11 months. To the telescopic observer this 

 star does not actually disappear, but when it is faintest it gives only 

 about one per cent of the visual light which it gives when it is 

 brightest. Since the time of Fabricius and Holwarda, the number of 

 recognized stars of this type has increased, especially during the last 

 few years, until now our catalogues contain about two thousand 

 other examples. 



All of the stars of this type are more or less red and appear to be 

 among the coolest stars in the sky. In their outer layers at least, 

 they apparently vary in temperature and in color as well as in 

 luminosity, but the highest temperature exhibited by a variable of 

 this type never exceeds that of a very hot furnace, while many of 

 the blue stars show temperatures ten times as high. Among the 

 giant stars, in general, low temperatures and large diameters are 

 very closely correlated and are thought to be characteristic of the 

 earliest stages of a star's development. The enormous size of Mira 

 is illustrative, probably, of all long-period variable stars. If the 

 sun were placed at the center, inside the body of Mira, and the dis- 

 tance (92,900,000 miles) between the sun and its companion the 

 earth were unchanged, the earth would be several million miles be- 

 low the surface of the star. Not only the earth, but Mars, which is 

 50 per cent farther away from the sun than the earth, would be 

 beneath the surface of this giant star. 



At the level which we call the surface of IMira, the constituent 

 gases are much rarer than those composing the atmosphere of the 

 earth, 50 miles above its surface; in fact, they are in a rarer condi- 

 tion than can be duplicated by exhausting the air of a vessel with the 

 best air-pump yet constructed. The rarity of the gases at the sur- 

 face of this star is due to the smallness of the gravitational pull 

 which varies inversely as the square of the diameter. A man weigh- 

 ing 200 pounds on the earth would weigh less than 4 ounces at the 

 surface of Mira, were it possible to weigh him with spring balances 

 in each case. On the small dwarf star circling about Sirius, by way 

 of further contrast, this man weighed by the same method would 

 weigh no less than 3,000 tons, w^hereas on our sun he would weigh 

 only 5,500 pounds, or someAvhat less than 3 tons ! 



In many ways the class of variable stars to which Mira belongs, 

 ordinarily called long-period variables, is closely related to another 

 type known to astronomers as Cepheid variables. The earliest ob- 

 served type of this class is Delta Cephei, discovered by Goodricke 

 in 1784. These stars are not so red as the long-period variables, and 

 their periods of variation extend from a few hours to about 70 days, 

 whereas the periods of long-period variables range from about 100 



