RECORDS 31 



we do about blood corpuscles. By actual count it is known 

 that there are four to six millions of the latter in a cubic milli- 

 meter; and with equal definiteness calculation shows us that 

 there are about a million million million molecules in a cubic 

 millimeter of the air around us. Notwithstanding this appar- 

 ently crowded assemblage, the individual molecules move about 

 in the liveliest manner, their average speed being about five hun- 

 dred meters per second, and this in spite of the fact that the 

 average length of an unimpeded journey is barely visible by the 

 aid of the best microscopes. Each molecule must therefore 

 collide with its neighbors astonishingly often, the encounters 

 occurring, in fact, about five thousand million times per second.^ 



More surprising still than the properties of assemblages of 

 molecules forming gases are the properties of the individual 

 molecules, especially when they are made up of two or more 

 atoms. Such miniature systems, comparable, probably, in com- 

 plexity with the Martian and Jovian subsystems of the solar 

 system, exhibit degrees of constancy which rival the invariable- 

 ness of the fixed stars themselves. This is particularly the 

 case with their rates of vibration as disclosed by the spectro- 

 scope. These rates afford one of the most delicate tests of the 

 properties of matter, whether it is found on the earth or on the 

 most distant star ; and yet the vibrations, which recur with a 

 regularity equal to, if not surpassing, the regularity of the rota- 

 tion of the earth, are executed at the rate of some hundreds of 

 millions of millions per second.^ Herein, perhaps, we may find 

 a cosmic unit of time as well as a cosmic unit of distance, 

 though both appear to be inconveniently small for terrestrial 

 purposes. 



But the smaller bodies of the universe do not end with mole- 



1 See, for example, " The Kinetic Theory of Gases," by Dr. Oskar Emil Meyer, 

 translated by Robert E. Baynes, Longmans, Green & Co., New York, 1899. 



^ The number of vibrations per second corresponding to any given wave-length 

 of light may be easily computed. For the velocity of light is about 300,000 kilo- 

 meters, or 3 X 10'* microns per second, and this divided by the wave-length in 

 question gives the number of vibrations per second. Thus the average wave-length 

 of the cadmium rays used by Professor Michelson (cited above) is about half a mi- 

 cron. The material sources of these rays must vibrate, therefore, about six hun- 

 dred million million times per second. 



