PHYSICS — COBLENTZ. 25 1 



CobIent2, William W"., Bureau of Standards, Washington, District 

 of Columbia. Grant No. 198. Contimiatio?i of investigatio7i of 

 infra-red absorption and emission spectra. 5^1,000. 



Report. — The experimental part of this work, which has occupied 

 two years from June, 1903, was performed at the physical laboratory 

 of Cornell University. The problem was to determine the effect of 

 molecular weight upon absorption ; also the effect of chemical struc- 

 ture, i. e., the arrangement of the atoms in the molecule, and the 

 effect produced b}^ the substitution of a CH^ or OH group of atoms. 

 The problem is long and intricate and only a few of the main results 

 obtained will be summarized here : 



(i) The infra-red absorption spectra of some 135 compounds, 

 mostly of hydrogen and carbon, have been investigated to 15M, using 

 a mirror-spectrometer, a rock-salt prism, and a Nichols radiometer. 

 The compounds include solids, liquids, and gases. 



(2) A comprehensive study of isometric compounds shows that the 

 arrangement or bonding of the atoms in the molecule, i. c., its struc- 

 ture, has a great influence upon the resulting absorption spectrum. 



(3) The maxima of absorption do not shift with increase in molec- 

 ular weight. 



(4) A rise in temperature of 20° has no effect upon the trans- 

 parency of the compound, nor upon the position of its maxima of 

 absorption. 



(5) Total absorption is not influenced by the size of the molecule, 

 while compounds having sulphur or halogens are more transparent 

 than those having H, O, OH, or N which they replaced. 



(6) The spectra of groups of compounds are similar and are char- 

 acteristic of the grouping adopted by chemists. 



(7) Carbohydrates have a characteristic spectrum with well-defined 

 absorption bands in the region of 0.85, 1.7,3.4, 6.8, and 13.8/^. Cer- 

 tain bands are closely harmonic. 



Infra-red emission spectra. — The investigation consisted of two 

 parts, viz, emission spectra of metals or salts of the metals in the 

 electric arc, at atmospheric pressure, and the emission spectra of 

 gases in vacuum tubes. 



Such an investigation is of twofold interest : (i) Previous exper- 

 iments showed that there are strong emission lines just beyond the 

 red, which became less intense farther toward the infra-red (1.2 Ai), 

 and the question arose whether emission lines exist beyond this point ; 

 (2) theoretically, all the emission lines predicted by our spectral 

 series formulas end just beyond the red — about 2 ,«. These formulae 



