of the Heat-conducting Power' of Mercury* 485 



A 



Herein ^ is still to be determined, which is done as follows. 



In any cross section near the end of the rod, where only current- 

 action takes place, we have 



* &(l.+ &)=ph{1>-J0), 



or 



therefore 



2 



A t'-0 



If this value be inserted and the above equation be given a sim- 

 pler form, it is readily converted into 



l+/^/T-/'y 



(1 + aT) 2 



In accordance with this formula, therefore, we have in the ex- 

 periment, besides the surrounding temperature, only the three 

 temperatures T, t r , and m to determine and thus obtain a number 

 of equations between « and ft. 



§3. 



To this purpose I now directed the following apparatus for 

 the investigation of mercury. On the proper experimental tube 

 (filled with mercury which had been during several months treated 

 with protonitrate of mercury and thereby rendered very pure), 

 of about 1 centim. diameter, and nearly 60 centims. length, and 

 very carefully selected as accurately cylindrical, only the tempe- 

 ratures T and m were to be determined. A mercury thermo- 

 meter extending to 200° C, and with a very small cylindrical 

 bulb, served for the determination of T. It was ground into a 

 small glass lateral appendix borne by the tube immediately below 

 its upper extremity, and with its bulb occupied only a small 

 portion of the tube's diameter. The indications of this thermo- 

 meter of course correspond to the mean temperature of the mer- 

 cury in its bulb. Now, since this mercury itself, and indeed 

 perpendicular to the axis of the cylinder of the bulb of the ther- 

 mometer, took part in the heat-conduction, the mean tempera- 

 ture indicated existed in a cross section through the thermo- 

 meter-bulb and the experimental tube the course of which ran 

 in the upper half of the bulb. If, then, in the above calculation 

 we put for the upper temperature of the entire mercury- tube 

 that indicated by the thermometer, we have had in reality, with 



