ON THE INTENSITY OF SOLAR RADIATION. 145 



difference in the readings obtained with a thermometer of ordinary white 

 glass. 



On May 22 three sets of observations were made, two with the green, 

 glass and one with the white-glass thermometer : those with the green 

 were made between X.17 and X.50 and between XI.35 and XI.53, 

 that with the white-glass instrument between XI. and XI.25. 



The observed excesses of temperature of the green-glass thermometer 

 above the temperature of the case were 48°'3 F. and 49°"3. The observed 

 excess of the white glass was only 32°'8. The corresponding calculated 

 excesses obtained by the method described in the last Report were 

 respectively 50°-29, 49°-24, and 33°-30. 



It is thus seen that the white- glass bulb rises to about two-thirds .of 

 the excess indicated by the green-glass bulb. This, however, is no dis- 

 advantage, for when the temperature of the insolation thermometer is 

 much above that of the case the simple law which for smaller excesses 

 connects the rate of cooling with the difference of temperatures is no 

 longer a sufiBciently near approximation, and the reduction of the observed 

 results becomes more difl&cult. 



As the simultaneous reading of the three thermometers is not an easy 

 operation, an attempt has been made to replace them by two thermo- 

 electric junctions. A copper disc, 20 mm. in diameter and about "75 mm. 

 thick, was soldered at its centre to a piece of iron wire. The wire was so 

 bent that when the centre of the disc opposite to the soldered joint is 

 exactly behind the hole in the copper cube, the other end of the wire 

 makes contact with the copper cube midway between the front and back. 

 To the edge of the disc a thin copper wire is soldered, which passes 

 through a glass tube in the central opening of the cube, and is thus insu- 

 lated from it. The experiment being only preliminary, the iron wire has 

 been fixed in a hole drilled in the copper plug which usually holds the 

 insolation thermometer, the glass tube carrying the insulated wire being 

 passed through the hole in the same plug. The other terminal from the 

 copper cube is made by fixing a piece of copper wire in the plug which 

 closes the hole of the case thermometer B in front of the cube. In a per- 

 manent instrument a binding screw should be attached to the cube in the 

 plane of the disc. To increase the absorption of heat by the copper disc, 

 it was blackened by being placed for a short time in sulphuretted 

 hydrogen. The black surface thus obtained does not, however, com- 

 pletely absorb the radiation, for, on throwing a beam of sunlight on it, it 

 is observed that some of the light is scattered. The surface thus obtained 

 may, in addition, be not permanent. 



The terminals of the thermo-couples were connected to a reflecting 

 galvanometer of "97 ohm resistance, and the disc exposed to the raj's of 

 the sun, the lens of the instrument being used. The deflection of the 

 galvanometer became steady after an exposure of from five to eight 

 minutes, whereas twenty minutes were required when the green-glass- 

 bulb thermometer was used. 



In order to determine the value of the deflections a double thermo- 

 couple was made by soldering to two stout copper wires a bent piece of 

 thick iron wire. Close to the junctions delicate thermometers were tied, 

 and the apparatus was so arranged that the thermo- junctions and thermo- 

 meter bulbs could be plunged in test tubes containing paraffin oil : one 

 of these test tubes could be heated, and the connections were so made 

 that the current produced by the heated junction opposed that from the 



