116 M. E. Pringsheim on the Radiometer \ 



a gradual rise of temperature, and, in consequence of it, only 

 a gradual deflection of the vane. If now, after the apparatus 

 has, under the influence of the irradiation, taken a fixed position 

 of rest (in which therefore the rotation-force generated by the 

 excess of temperature of the front over the back side is kept 

 in equilibrium by the torsion of the glass thread), the irradia- 

 tion be suddenly interrupted, the back side of the lampblack 

 suddenly loses the heat conveyed to it by absorption every 

 instant during the irradiation, while the heat communicated 

 to the mica face by conduction diminishes just as slowly as it 

 increased at the beginning. Hence at the first instant of the 

 shading the temperature-difference of the two faces is increased 

 by the same quantity, although in the inverse direction, as at 

 the first moment of the irradiation ; and the increase produces 

 an equally strong and sudden deflection of the vane as in the 

 first case, only in the opposite direction. When then from 

 the sudden cooling of the front layer of the lampblack a gra- 

 dual lowering of temperature takes place at the front face of 

 the mica, the difference between the two sides gradually comes 

 to an end, and the vane slowly returns to its initial resting- 

 position. 



That this pregnant effect of heat-conduction is not brought 

 to light with the usual mica radiometers is moreover a clear 

 proof of their unsuitableness when the object is to fathom the 

 causes of their motion. If, namely, such a radiometer is irra- 

 diated on one side so that the clear faces of the vanes is always 

 illuminated and the dark sides are in the shade, the first de- 

 flection, produced by the absorption of the back side of the 

 lampblack, is in every case sufficient to carry the irradiated 

 vanes completely out of the range of the radiation and bring 

 a fresh vane into it, which in its turn, by its first deflection, 

 continues the rotation of the radiometer- cross in the same 

 direction, so that a constant rotation, with the mica going in 

 front of the lampblack, sets in, and the effect of the heat- 

 conduction completely vanishes. 



Therefore, although an effect of the heat-conduction had 

 unequivocally come out already in the experiments with lamp- 

 blacked mica vanes, yet corresponding experiments were 

 undertaken with a very well conducting substance, in order 

 to see whether the phenomena would be essentially affected by 

 the good conductivity of the vane. For this purpose a vane 

 of very thin copper foil (so-called Dutch metal), an excellent 

 conductor of heat, was prepared and introduced into the bulb. 

 First some experiments corresponding with the previously 

 communicated observations with the translucent mica vane, 

 were instituted with the vane not lampblacked, but bright on 

 both sides. 



