192 ANNUAL, REPORT SMITHSONIAN INSTITUTION, 19 3 9 



intermediaries between atoms and organisms on our scale provide a 

 method of approach. Also in a few instances it is possible to observe 

 visually the effects produced by very rapidly moving atoms. Our 

 chief quantitative information, however, is received very indirectly 

 but unambiguously through the refined techniques of electrical 

 measurements and consists of a series of pointer readings. The 

 information that we obtained is a rough estimate of atomic size, more 

 accurate information about the forces that hold atoms together and 

 that atoms exert on one another, and very precise information as to 

 atomic masses or weights. The result is that we know atoms are of 

 such a size that about a hundred million of them laid side by side 

 would extend across your finger nail, the smallest weight that you 

 can feel in your hand is a million millions times greater than the 

 force holding two atoms together, and about 24 zeros would have 

 to be written before the figure representing the mass of a thimbleful 

 of water before it would represent the mass of an atom. The 

 determination of an atomic mass is representative of the complex 

 indirect methods of measuring atomic quantities. In the first place, 

 it must be known that an atom can be given an electric charge. 

 Then we must know the magnitude of this charge. We must also 

 have observed that a rapidly moving charged atom is deflected from 

 a straight path into a curved one by a magnetic field. We must have 

 a theory that tells us a relation between the velocity of the atom 

 and an observable quantity such as a voltmeter reading, and one that 

 relates the known charge, radius of curvature, magnetic field, and 

 velocity with the unknown mass. Then from a voltmeter reading, a 

 meter reading giving the magnitude of the magTietic field, and a 

 length measurement that gives us the radius of curvature we are able 

 to measure the atomic mass with an accuracy determined by the least 

 accurate of all the measurements that have gone into a determina- 

 tion of this quantity. In this way we can find the masses of all 

 the different kinds of atoms with an accuracy of about 1 percent. 

 However, relative masses, that is, the ratio of the masses of two 

 different types of atoms, can be determined much more exactly, as 

 these relative measurements can be made to depend, for instance, on 

 the ratio of the two lengths involved in measuring the radii of 

 curvature, and they can be found to 1 part in 100,000. This general 

 fact is true throughout all these types of measurements; absolute 

 measurements in terms of length, mass, and time are difficult to make 

 accurately, but in many instances ratios are of the greatest importance 

 and these can in general be found with much greater precision. 



The next question of interest is what structure if any these minute 

 entities possess. In posing this question we must realize that we 

 can never answer it in the detail that we can, for instance, specify 

 the structure of a watch or a locomotive. In the first place we learned 



