Prof. Wartmann on the Polarization of Atmospheric Heat. 109 



columns of air. The variations in the transparency of the air, 

 the calorific reflexions which proceed from the surface of the 

 earth and from the clouds, render in general the intensity of the 

 heat radiating in any given direction extremely inconstant. 



These obstacles being known, I endeavoured to combat them 

 by the following arrangements : — The thermo-electric pile of 

 Melloni was placed in a capacious chest, so that its uncovered 

 face was turned towards an opening in the centre of one of the 

 sides. This face is provided with its cone of polished brass 

 fixed in a cylinder of wood, which is lined with a tube of paste- 

 board. The extremity of this tube enters the circular opening 

 in the side, and moves in it with strong friction ; screens and 

 diaphragms of various substances can be attached to it. At its 

 extremity, the piece destined to contain the analyser is fixed level. 

 It carries a collar, to which the hand imparts a rotative motion 

 by means of a strong handle, and which carries an index pointing 

 to a dial, three decimetres in diameter, fixed against the chest. 

 This piece is surrounded by a cylindrical case of white paste- 

 board blackened in the interior, six decimetres long, open in 

 front, and destined to circumscribe the portion of space to be 

 examined, the oblique rays being arrested. 



The analyser which I made use of in my first experiments was 

 a pile of thin plates of mica. It would have been easy to render 

 it moveable round a line perpendicular to the axis of the thick 

 pasteboard cylinder which enclosed it ; but I preferred arresting it 

 at an angle of 35° with this axis, and placed a similar pile parallel 

 to it and six centimetres in advance. This assemblage polarizes 

 and analyses the heat completely ; it prevents the currents of 

 air from acting upon the solders of bismuth and antimony, and 

 destroys the radiation of those metals so effectually as to render 

 all other preservative unnecessary. I afterwards replaced this 

 portion of the instrument by a very large Nicol's prism con- 

 structed by M. Ruhmkorff. It is 0*086 of a metre in length ; 

 the greater diagonal of the base is 0-036 of a metre, and the 

 lesser 0*028 of a metre*. 



* With this apparatus I have repeated the experiments which I published 

 in 184G, relative to the rotation of the plane of polarization of radiant heat 

 jmder the influence of magnetism. A solar ray traverses a Nicol's prism 

 0O7 of a metre long, the diagonals of which measure 0-03 and 0023 of 11 

 metre respectively. It then passes through a parallelopiped of heavy glass 

 0'029 of a metre long, and the square hases of which measure 001/5 of a 

 metre the side. This glass is made by Prof. Faraday, to whose kindness 1 am 

 indebted for it. It was placed between the polar pieces of an clectro-inag- 

 n.-t, the soft iron cylinder! of which were 009 of a metre in diameter, and 

 earned nine layers of copper wire 0003 of a metre thick, each layer con- 

 ■uting ol'sixtv winding!. This wire is brought into the circuit of a battery 

 of ten of Grove's large cells. The ray arrives at the thermo-electric pile 



