Let the temperature of a dark room be observed by means of a very sensible air-thermometer; if the bulb is then brought to the direct sun light as it shines into the room, nothing is more natural than that the fluid should indicate a much higher degree of warmth. If upon this we interpose coloured glasses, it follows again quite naturally that the degree of warmth must be lowered; first, because the operation of the direct light is already somewhat impeded by the glass, and again, more especially, because a coloured glass, as a dark medium, admits less light through it.
675.
But here a difference in the excitation of warmth exhibits itself to the attentive observer, according to the colour of the glass. The yellow and the yellow-red glasses produce a higher temperature than the blue and blue-red, the difference being considerable.
676.
This experiment may be made with the prismatic spectrum. The temperature of the room being first remarked on