84 ANNUAL RECORD OF SCIENCE AND INDUSTRY. 



portion of air which is not equally dense on all sides of the 

 ray, it is deflected toward the side on which the density is 

 greatest, the sharpness of the curvature being proportional 

 to the rates at which the density varies. If the air is strati- 

 fied horizontally, it follows that a ray traveling nearly hori- 

 zontally will be bent the most, and it is by such rays that 

 we see the images which constitute mirage. In the average 

 state of the atmosphere the curvature of horizontal rays is 

 about one fifth or one sixth of the curvature of the earth's 

 surface, being greater in cold than in warm weather, and 

 greater with high than with low barometer. The curvature, 

 however, depends principally upon the rate at which the 

 temperature changes with the height. The average rate is 

 one degree for 300 feet. If the rate were one degree for 

 fifty-three feet, the horizontal rays of light would be straight 

 lines. A more rapid rate than this will render the air above 

 denser than that below, and cause rays to bend up instead 

 of down. This condition of affairs may exist for a time, 

 although it is a condition of unstable equilibrium, and must 

 eventually be broken up by the inflow from the surround- 

 ing regions of cold air. An increase of temperature up- 

 ward at the rate of about one degree in sixteen feet will 

 make the curvature of horizontal rays equal to that of the 

 earth, so that they may encircle the globe. Any such 

 downward bending of rays of light increases the range of 

 our vision, enabling us to see around the horizon, which 

 otherwise limits the view, thus bringing distant objects 

 in sight, and rendering nearer objects more distinctly vis- 

 ible, but without in any way inverting them. 12 A, XL, 

 151. 



THE NEW SELF-RECORDING BAROMETER. 



A self-recording barometer has recently been made by Mr. 

 Kedier, which seems to have many excellent points, and to 

 be enthusiastically received in France. In a communication 

 to the Meteorological Society of Paris, he states that his in- 

 strument consists of an ordinary syphon barometer carrying 

 a very light ivory float, upon which is fixed a vertical steel 

 wire terminating in a point. A horizontal needle rests upon 

 this point, its other extremity is in connection with a double 

 series of clock-work, the wheels of which move either forward 



