STRUCTUEE OF THE GALAXY — PLASKETT 199 



have is in favor of such a conception, while the essential similarity 

 in structure of all the spirals forms strong presumptive evidence of 

 a similar structure in the galaxy. That the galaxy is essentially a 

 homogeneous unit like the spirals is indicated by the demonstration, 

 based mainly on observational data obtained at Victoria, that the 

 space between the stars is uniformly pervaded throughout by almost 

 incredibly diffuse matter, millions of times more tenuous than the 

 highest vacuum we can produce on the earth, with the molecules 

 about a centimeter apart, and so tenuous that there would be only 

 about 4 ounces of this gaseous matter in the whole volume of the 

 earth. Our conception of the galaxy as a single homogeneous dy- 

 namical unit is much strengthened by the proof to be presently given 

 that both stars and diffuse matter are in beautifully ordered rotation 

 in their own plane. 



Another view of the galaxy was, however, advanced about a year 

 ago by Shapley, who was the originator of the present conception of 

 its dimensions and structure. Shapley contends that, as the galaxy 

 is five times larger than any other known system, it is not likely to 

 have a similar structure. He analyzes loose clusters of nebulae in 

 Virgo and Centaurus and assumes that the galaxy is built up of a 

 number of discrete nebulae loosely aggregated together. There are, 

 however, many difficulties in this hypothesis, for in the clusters he 

 assumes as analogous there is no trace of discoidal form, while the 

 component nebulae are widely separated from one another in con- 

 tradistinction to the galaxy, where star clouds merge into one an- 

 other. But the main objection to Shapley's new hypothesis lies in 

 the demonstrated dynamical unity of the galaxy and in its ordered 

 rotation in its own plane, to which his loose groups of nebulae have 

 not the least similarity. 



II. THE ROTATION OF THE GALAXY 



The conception we have tried to develop of the structure of the 

 galax}^ should enable us to understand better the motions in the 

 galaxy. The name, the " fixed stars ", is a misnomer, as it has long 

 been known that motion is a universal property of the stars. The 

 motion of a star, which may be in any direction, can only be meas- 

 ured in two directions — its cross motion, or change of position in 

 the sky, called the proper motion, and its motion in the line of sight, 

 its radial velocity measured by the spectroscope. From these two 

 components, provided the distance is known, we can get the real or 

 space motion of any star. The proper motions are all very small, 

 the largest being a change of position of 10" per year, or the width 

 of the moon in 180 years, while the majority of proper motions are 

 less than one-hundredth of this, or moon width in 18,000 years. The 



