188 ANNUAL REPORT SMITHSONIAN" INSTITUTION, 19 2 8 



mum to maximum much more quickly than they fall from maximum 

 to minimum. The variation is usually a little less than one magni- 

 tude, the maximum brightness being about twice that at minimum. 

 The spectroscope shows that the surface is heaving up and down with 

 the period of the star's variation and that changes of temperature 

 are taking place. Due to some unknown cause, the gaseous body of 

 a star seems to have taken up a periodic pulsation of some kind, 

 which accounts for the action observed there. If we assume with 

 Shapley and Eddington that this pulsation is a free oscillation, the 

 mass and density of the star may be computed. The period will de- 

 pend largely on the density. Since the period may be observed with 

 extreme precision, changes in the density, if present, might be de- 

 duced by investigating secular changes in the period. Such changes 

 have not been definitely proved in the span of time during which 

 these stars have been observed, which gives us an idea of the extreme 

 deliberateness of stellar evolutionary processes. 



Again, the periods have been found to be correlated to the lumi- 

 nosities, Cepheids with the longer periods having the greater intrinsic 

 brightness. Wlien proper calibration has been made, the lumi- 

 nosities and distances of Cepheids can thus be determined provided 

 their periods are know'n. The great distances of the spiral nebulse 

 found by Hubble by this method are most interesting. 



Some 300 stars have this general type of variation. The periods 

 group about one-half day and 10 days. The longest known periods 

 of this type are about 50 days. Those with periods near one-half day 

 in length seem to form a separate group. They are of interest be- 

 cause they have the highest velocities of any stars known in oiir 

 Milky Way system. One of them, called VX Herculis, is moving at 

 a rate of over 250 miles a second, a speed sufficient to take it the whole 

 distance from the earth to the sun in 4 days' time. 



It may be doubted whether there are any stars wliich are entirely 

 constant in brightness. Variation seems to be a relative matter. 

 Nearly all very red stars appear to vary to a small extent in an 

 irregular way. Our sun shows small changes in intensity of about 

 2 per cent, which seem, in general, to follow the sun-spot activity in 

 an 11-year period, but with considerable irregularity. At certain 

 times there are also fluctuations within a few days. The spots them- 

 selves, the faculse, and the jironiinences are continually changing, 

 and all affect, in some degree, the total amount of radiation given 

 out. It is safe to say that activities of a similar nature could be 

 found in the atniosplieres of all stars, if we could get within obsei'v- 

 ing distance; but they are i)robably more characteristic of dwarf 

 stars of high density like our sun than of the tenuous giants. 



The sun is the only star on which we can actually observe details 

 of atmospheric activity. On account of its proximity we see its 



