THE RAXGE OF NATURe's OPERATIONS. 211 



4. That the number of molecules in a cubic centimeter of gas at 

 standard temperature and pressure is somewhere in the neighborhood 

 of a uno-twentyone. This follows as a corollary from (2). 



5. That the number of chemical atoms in a cubic centimeter of a 

 solid or liquid is a number of the same order as a uno-twentyfour. 

 This follows from (3). 



6. That the masses of the chemical atoms probably lie between the 

 twentysecondet and the twentytifthet of a gram. This follows from 

 (4) and from the known densities of solids and liquids. 



The tenth-metret, the smallest of the above measures, is the ten-thou- 

 sand-millionth part of a meter. It is about the two-thousandth part 

 of the smallest interval which the best microscope can detect when 

 most carefully handled. 



Another branch of physical inquiry' has introduced us into the same 

 region of magnitudes, and has even carried us farther. The wave- 

 lengths of visible light range from 38 to 76 eighth-metrets, and can, 

 by methods which will be described farther on, be measured with such 

 marvelous precision that it is possible to detect differences of wave- 

 length which amount to a very small fraction of a tenth-metret. 



nature's operations on a large scale. 



When we turn our attention to nature's operations on the large 

 scale, we find that the greatest lengths we can as yet succeed in meas- 

 uring are the distances of those few stars which have perceptible 

 parallax.^ The distances of these stars from the solar system range 

 from four to fifteen metro-sixteens. and it is not likely that any star 

 could send us light enough to be visible in any of our telescopes if a 

 thousand times more remote. At a distance, then, of about 10,000 

 metro-sixteens — that is, at a distance of about a metro-twent}" — our 



the range implied by the phrase "of the same order as" becomes widened. It miw 

 ranges from V'lOOO times the assigned value (in this case a mio-eighteen per cubic 

 millim. ) to 1 v/lOOO times this value; so that it includes 30, 20, 10 times, and 1/10, 

 1 20, and 1/30 of a uno-eighteen. Any of these numbers is much better represented 

 by a imo-eighteen than it Avould be by a uno-fifteeu, the number which is a thousand 

 times smaller, or by a uno-twenty-one, the number which is a thfnisand times larger. 

 The knowledge thus reached as to the number of molecules that are present may 

 seem very indefinite; but it is far from being valueless. 



'Attempts have been made to infer the parallax of binary systems from a s|)ectro- 

 scopic determination of the difference of velocity in the line of sight of the constitu- 

 ent stars, combined with the known periodic time and the apparent angular size and 

 form of the system. This method has been applied to y Virginis and to y Leonis 

 with results which are not yet free from doubt on account of the extreme delicacy of 

 the observations, but which seem to place these stars at distances, in the case of y 

 Virginis, of about 60 metro-sixteens, and in the case of y I>eonis of 150. These are 

 distances which are one step of our scale farther — i. e., about ten times farther — from 

 us than those of which the parallax can l)e directly measured. (See Astr. Xach. Xo. 

 3510, or Nature for August 25, 1898.) 



