of Flame and Gases. 409 



three at once; one over the axial line and one at each side. 

 When they were under the poles, the lower end was turned 

 up a little for the purpose of facilitating observation there. 



The gas delivered at the poles, as already described, con- 

 tained a little muriatic acid (obtained from the solution in the 

 paper), but not enough to render it visible. To make it ma- 

 nifest up which catch-tube it passed, a little piece of bibulous 

 paper, folded and bound round and suspended by a copper 

 wire, was dipped in the solution of ammonia and hung in each 

 of the tubes. It was then evident at once, by the visible fume 

 formed at the top of one of the tubes, whether the gas delivered 

 below passed up the one or the other tube, and which : and 

 yet the gas was perfectly clear and transparent as it passed by 

 the place of magnetic action. 



In addition to these arrangements, I built up a sheltering 

 chamber about the magnetic poles and field, to preserve the 

 air undisturbed. This was about six inches long by four inches 

 in width and height, and was easily made of thin plates of 

 mica, which were put together or taken down in a moment. 

 The chamber was frequently left more or less open at the top 

 or bottom for the escape of gases, or the place of the catch- 

 tubes. Its advantages were very great. 



yJir. — In the first place air was sent in under these arrange- 

 ments, the stream being directed by the axial line. It made 

 itself visible in the catch-tube above by the smoke produced; 

 but whether the magnet was active or not, its course was the 

 same; showing that, so far, the apparatus worked well, and 

 did not of itself cause any erroneous indications. 



Nitrogen. — This gas was sent from below upwards, and 

 passed directly by the axial line into the catch-tube above; 

 but when the magnet was made active, the stream was affected, 

 and though not stopped in the middle catch-tube, part ap- 

 peared in the side tubes. The jet was then arranged a little 

 on one side of the axial line, so that, without the magnetic 

 action, it still ascended and went up the middle catch-tube: 

 then, when the magnetic action was brought on, it was clearly 

 affected, and a great portion of it was sent to the side catch- 

 tube. The nitrogen was, in fact, manifestly diamagnetic in 

 relation to common air, when both were at the same tem- 

 perature; but as four-fifths of the atmosphere consists of ni- 

 trogen, it seemed very evident, from the result, that nitrogen 

 and oxygen must be very different from each other in their 

 magnetic relations. 



Oxygen. — A stream of oxygen was sent down through air 

 between the poles. When there was no magnetic action it 

 descended vertically, and when the magnetic action was on it 



