The Evolution of Chemistry. 135 



methods of crystallization of salts, oxides and chlorides 

 formed, etc., the hypothetic and real elements agree exactly. 

 The predicted atomic weight of gallium was 69 and the 

 found weight was 69. The predicted atomic weight of 

 scandium was 44 and the found weight was 44. In the 

 presence of such facts is it not strange that there are in- 

 telligent men who pretend to believe that atoms are as un- 

 real and intangible as hobgoblins and fairies? So really 

 physical, indeed, are the molecules made up from these 

 atoms that Sir William Thomson and others have been able 

 to calculate their approximate weights in terms of fractions 

 of a grain. 



No one can carefully study the periodic law of Mendel- 

 ejeff, comparing it at the same time with a homologous 

 series and its heterologous derivatives, without being struck 

 with the idea that the atoms are products of evolution. If 

 all the properties of matter are simply due to the weight of 

 the little pieces from which it is built (and so the law de- 

 clares), then at bottom every element must be the same. 

 This implication is confirmed by spectrum analysis. 



Within the present generation Prof. Bunsen and Prof. 

 Kirchhoff, of Heidelberg, Germany, devised a plan b}~ which 

 the composition of sun and stars might be accurately de- 

 termined from their light. The instrument used is made 

 of prisms that separate the different colors found in the 

 beam to be examined. Certain lines of colored light are 

 given forth by every element, and a knowledge of the ap- 

 pearance and places of these lines enables one to tell just 

 what element is coloring a flame. If such light passes 

 through a vapor of the same element before reaching the 

 prism, dark bars appear just where the colored lines should 

 have been that belonged to that element. By the initiated 

 the bars are as easily read as the colored lines. In ordinary 

 chemical analysis one one-hundred-and-twentieth of a grain 

 approaches very nearly the lowest limit of practical determi- 

 nation. The spectroscope, however, is so sensitive that it 

 can tell the presence of a substance when the quantity is 

 nearly two million times less than this, or one two-hundred- 

 and-forty-millionth of a grain. 



Very soon after the spectroscope was sufficiently perfected 

 for practical work the four elements ca?sium, rubidium, 

 thallium, and indium were discovered by its aid. The 

 workers knew they were there from the lines they gave, al- 

 though they had never been isolated. The use of this in- 



