Clifton College Scientific Society. 9 



exclude daylight, and is kept by means of a clock slowly re- 

 volving on one side of a barometer, whilst a gas-light is con- 

 stantly burning on the other side. The light cannot pass 

 through the mercury; so the only effect produced upon the 

 sensitized paper is from the light that passes through the space 

 between the surface of the mercury and the top of the tube. 

 This space is, of course, larger or smaller, according as the 

 mercury falls or rises ; and the black band on the paper will be 

 proportionally broader or narrower. Thus the variations of the 

 lower edge of this band become a faithful rejDresentation of the 

 changes of height in the mercurial column. There is a mechan- 

 ical contrivance connected with the clock, by which the light of 

 the gas is intercepted during four minutes, at every even hour, 

 thus leaving a white line on the photograph, which accurately 

 points out the hour. 



By means of a tabulating instrument, the barogram is measured 

 at every hour to the thousandth of an inch, and the record 

 entered on sheets ruled for the purpose. 



The changes of temperature in the dry and wet bulb ther- 

 mometers are also photographed, though in a rather different 

 manner. A vertical cylinder, covered with sensitized paper, as 

 mentioned above, is made to revolve by a clock, as has been 

 already described with reference to the barograph. The ther- 

 mometers are placed within a screen outside the building, about 

 three feet from the wall, and being bent at a right angle, they 

 come through the window-sill into the thermograph room. 

 Towards the top of the thermometer there is a small detached 

 column of mercury, separated from the rest by a little speck of 

 air, and as the temperature varies, the air-speck rises or falls 

 also. The gas burners in the room are reflected by highly 

 polished mirrors upon the upper portions of the thermo- 

 meters, and the light, passing only through this small inter- 

 stice formed by the air-speck, falls upon the sensitized paper, 

 and marks a line, more or less wavy, according to the changes 

 of temperature. The hours are marked on the curve by cutting 

 off the light in the manner already described; and the residts are 

 tabulated and recorded for every hour. 



Tlie thermometers commonly in use for ascertaining the 

 maximum and minimum temperatures, may give pretty accurate 

 results, but they do not, of course, record lolien these extremes 

 occurred. By means, however, of these continuous curves, we 

 are enabled also to fix the exact time of their occurrence. One 

 singular fact is thus made known to us — that the maximum 

 temperature is occasionally at or near midnight. Thus during 



