THE PERMIAN PERIOD. 673 



Primary independence in selective action. — It is important to note that in 

 selective action each gas absorbs its own wave-lengths, and these only. No mat- 

 ter what its quantity or its relative competency, it cannot do the selective work 

 of another gas. This is not experimentally demonstrable in the dark spectrum, 

 from lack of sufficiently refined means of analysis, but in the visible spectrum 

 the absorption lines of each substance differ from those of all other substances, 

 so far as perfect determinations can be made, and there is no reasonable ground 

 to doubt that this holds true in the dark spectrum also. Where bands appear 

 to overlap, they are to be interpreted as intermingled groups of lines which are 

 individually distinct. It is an error, therefore, to suppose, as some seem to 

 have done, that any amount of water- vapor can replace, or interfere with, the 

 absorption by carbon dioxide. Of the vibrations of various wave-lengths sent 

 forth by the earth, a certain set are absorbed by the molecules of water, a cer- 

 tain other set by the molecules of carbon dioxide, some few others by the other 

 gases, and the rest escape, except as caught by dust, clouds, general absorption, 

 etc. This relates to primary action, and to a certain portion of the secondary 

 action. 



After a molecule of carbon dioxide has absorbed its appropriate radiation, 

 and increased its vibratory energy correspondingly, it radiates wave-lengths 

 of its own kind, which can be taken up by other molecules of carbon dioxide, 

 but not by the molecules of other gases. The number of re-absorptions 

 and re-radiations of such wave-lengths is dependent upon the number of 

 dioxide molecules that lie in the paths of such re-radiations, which are in all 

 directions, some being, of course, back toward the earth, or laterally, where 

 they will be re-absorbed in adjacent parts of the atmosphere. The same 

 is true, of course, of all the other molecules relative to vibrations of their 

 own wave-lengths. Thus there is measurably perpetuated independence of 

 action. 



Thermal transference by molecular contact. — While, however, the preceding 

 action is in progress, each molecule is colliding with its neighbor molecules many 

 million times per second, and communicates some of its heat energy by such 

 contact. The molecules thus heated radiate wave-lengths of their own order, 

 and these may be absorbed by molecules of the same kind. Thus vibrations 

 may be transformed so as to be absorbed by all the constituents of the air in 

 succession, and their retention within the air be greatly prolonged, because of 

 the great mass of some of these constituents. 



The critical point is obviously the first absorption that starts the transfer 

 by impact. If there is no constituent to take up the vibration from the ether, 

 it escapes, but if it, is once taken up by a molecule, it may be transferred a mul- 

 titude of times before it escapes. 



Just what is the measure of the efficiency of carbon dioxide in contributing 

 to the temperature of the earth is not yet determined, and so the influences of 

 changes in its amount on the climate are as yet undeterminable, but as one of 

 the factors in the complex of agencies, it seems to require a place in our working 

 hypotheses. 



