26 FRAGMENTS OF SCIENCE. 
mutual attraction, and under suitable circumstances they 
coalesce to form compounds. Tims oxygen and hydrogen 
are elements when separate, or merely mixed, but they 
may be made to combine so as to form molecules, each con- 
sisting of two atoms of hydrogen and one of oxygen. In 
this condition they constitute water. So also chlorine and 
sodium are elements, the former a pungent gas, the latter 
a soft metal; and they unite together to form chloride of 
sodium or common salt. In the same way the element 
nitrogen combines with hydrogen, in the proportion of 
one atom of the former to three of the latter, to form am- 
monia. Picturing in imagination the atoms of elementary 
bodies as little spheres, the molecules of compound bodies 
must be pictured as groups of such spheres. This is the 
atomic theory as Dalton conceived it. Now if this theory 
have any foundation in fact, and if the theory of an ether 
pervading space, and constituting the vehicle of atomic 
motion, be founded in fact, it is surely of interest to ex- 
amine whether the vibrations of elementary bodies are 
modified by the act of combination whether as regards 
radiation and absorption, or, in other words, whether as 
regards the communication of motion to the ether, and the 
acceptance of motion from it, the deportment of the 
uncombined atoms will be different from that of the 
combined. 
4. Absorption of Radiant Heat by Gases. 
We have now to submit these considerations to the only 
test by which they can be tried, namely, that of experi- 
ment. An experiment is well defined as a question put to 
Nature; but, to avoid the risk of asking amiss, we ought 
to purify the question from all adjuncts which do not 
necessarily belong to it. Matter has been shown to be 
composed of elementary constituents, by the compounding 
of which all its varieties are produced. But, besides the 
chemical unions which they form, both elementary and 
compound bodies can unite in another and less intimate 
way. Gases and vapors aggregate to liquids and solids, 
without any change of their chemical nature. We do not 
yet know how the transmission of radiant heat may be 
affected by the entanglement due to cohesion; and, as our 
object now is to examine the influence of chemical union 
alone, we shall render our experiments more pure by liber- 
