of Flame and Gases. 413 



All the compounds of oxygen and nitrogen seem to show 

 the influence of the presence of the oxygen. Nitrous acid 

 seems to be less diamagnetic than air. Nitric oxide mingled 

 with nitrous acid and warm, is about as air. Nitrous oxide is 

 clearly diamagnetic in air, though it contains more oxygen: 

 but it also contains more nitrogen than air, and is also denser 

 than it, so that there is more matter present; still I think 

 the results are in favour of the idea that oxygen is diamagnetic. 

 By referring to the relation of carbonic oxide to carbonic acid, 

 described further on, it will be seen that the addition of oxygen 

 seems to make a body less diamagnetic. But the truth may 

 be, not that oxygen is really magnetic, but that a compound 

 body })Ossesses a specific diamagnetic force, which is not the 

 sum of the forces of its particles. 



It is very difficult to form more than a mere guess at the 

 relative degree of diamagnetic force possessed by different 

 gaseous bodies when they are examined only in air, because 

 of the many circumstances which tend to confuse the results. 

 First, there is the invisibility of the gas which deprives one of 

 the power of adjusting by sight so as to obtain the best effect : 

 then, there is the difference of gravity ; for if a gas ascend or 

 descend in a rapid stream, it may seem less deflected than 

 another flowing more slowly, though it be more diamagnetic; 

 and as to gases nearly of the specific gravity of air, whether 

 more or less diamagnetic, they are almost entirely dispersed in 

 different directions, so that little only enters the catch-tube. 

 Another modifying circumstance is the distance of the aperture 

 delivering gas from the axial line, which, to obtain the max- 

 imum effect, ought to vary with the gravity of the gases and 

 their diamagnetic force. Again, it is important that the mag- 

 netic field be not filled with the gas to be examined, and that 

 generally speaking only a moderate stream be employed ; which 

 however must depend again upon the specific gravity. 



The only correct way therefore of comparing two gases to- 

 gether is to experiment with them one in the other. For the 

 experiments made with gases, in gases or in air are differential, 

 and similar in their nature with those made on a formeroccasion 

 with solutions (Experimental Researches, 2362, &c.) ; I there- 

 fore changed the surrounding medium in a few experiments, 

 substituting other gases for air ; and first chose carbonic acid 

 as a body easy to experiment with, and one that would, pro- 

 bably, be more powerfully than some other of the gases, dia- 

 magnetic (I speak as to the appearances or relative results only) 

 in air. 



I constructed a kind of tray or box, by folding up a doubled 

 sheet of waxed paper; thus making a vessel thirteen inches 



