and on the Motion accompanying Heat-conduction. 167 

 Table III. — Radiating surface : — black glass. 



Radiating 

 surface. 



Analyzing 

 mirror. 



Series of experiments I. 







44-5 



46-25 



34-7 



35 



44 



44 5 



34 



34 



44 



43-5 



3475 



35 



45 



45 



34 



35 



44-5 



45-5 



35 



35-5 



44 



435 



3475 



34-5 



II. 



III. 



Horizontal 

 upwards. 



To the left. 



Horizontal 

 downwards. 



To the 



right. 



Left 



Horizontal 



Right 



Left 



Horizontal 



Right 



Left 



Horizontal 



Right 



Left 



Horizontal 

 Right 



45 



35 



44-5 



33 



44-25 



34 



45 



34 



44 



36-5 



44-25 



35 



Mean of all values with perpendicular planes of *) 

 reflection J 



Mean of all values with parallel planes of re- "1 

 flection J 



Total heat or sum 



Polarized portion or difference 



Percentage of polarized portion 



mean. 



45-2 



35 



44-3 



33-7 



44 



34-6 



45 



34-3 



44-7 



35-6 



44 



34-75 



mean. 



36 



27-7 



33-8 



28 



345 



26 



35-5 



27-5 



33-6 



27-5 



34-2 



26-8 



mean. 



421 



34-7 



45 



34-5 



42 



357 



42 



35-5 



45-2 



33-7 



42-3 



36-3 



44-8 



34-6 



79-4 

 10-2 

 12-9 



34-6 



27-2 



61-8 



7-4 



120 



431 



351 



78-2 



8-0 



10-3 



The agreement of these values with those given by the trans- 

 parent plate is a new proof that the transparent glass is no more 

 diathermanous for heat of 100° C. than the black opaque glass, 

 and that both kinds of glass possess radiating-powers for heat 

 which are very nearly equal. 



If the polarization of the radiated heat depended upon the 

 partial emergence of the heat from the interior and its refraction 

 on leaving the surface, then with a perfectly rough surface the 

 refraction must take place in all directions, and consequently no 

 polarization at all should be observable*. It appeared, there- 

 fore, of interest to try how the polarization was affected by using a 

 dull, ground glass plate in place of the polished one. The rough 

 glass plate was of the same kind of glass and of the same thick- 

 ness as the smooth one before used. As is seen from the follow- 

 ing numbers, the polarization was much less than with the 

 smooth plate ; but still polarization did occur. Hence the sug- 

 gestion immediately offered itself that the glass was not suffi- 

 ciently rough. Accordingly a perfectly rough substance, namely 



* " Wenn die Polarisation der ausgestrahlten Warme davon hemihrt, 

 dass die Strahlen zum Theil aus dem Innern kommen und beim Austreten 

 an der Oberflache gebrochen werden, so miisste fur den Fall, dass die Ober- 

 flache vollkommen rauh ware, die Brechung nach alien Richtungen stattfin- 

 den und daher gar keine Polarisation zu beobachten sein." As the above 

 appears ambiguous, I give the text literally. — F. G. 



