iilliiiillllllllllllllllluil n l illllilill mm) "VV 



286 Prof. Guthrie on the Conduction of Heat by Liquids. 



The accompanying figure shows the apparatus which I have 

 employed for showing, comparing, and approximately measuring 

 the conducting-powers of various liquids. 



A and B are two conical copper bottles 

 having very thin and perfectly flat bottoms 

 of exactly the same size. The two vessels 

 are fixed so that their flat surfaces are per- 

 fectly horizontal and at a little distance (say 

 two or three millimetres) from one another. 

 The liquid under examination is introduced, 

 by a pipette, between the two surfaces at W_, 

 where it remains, supported by adhesion. 

 The neck of the vessel A is fitted with two 

 tubes like those of a washbottle. The lower 

 vessel B serves as an air-thermometer, its 

 neck carrying a tube T which dips into a 

 reservoir of water B. Steam, or a current 

 of warm water, is passed through the ves- 

 sel A. By this means its lower surface is 

 raised to and maintained at a uniform tem- 

 perature. It is clear that as the liquid W 

 can only be heated from above, and as there 

 are no walls to the conducting column, any 

 expansion which is shown by the depression 

 of the water in T must arise from true con- 

 duction. 



It may be asserted that there is in all cases 

 a radiation through the liquid. If there be 

 a specific radiation independent of media, it 

 will clearly be a constant which may be sub- 

 tracted from the true conduction, and will 

 therefore not alter the order of conductive 

 power. If there be a radiation peculiar to 

 each medium, its manifestation may be fairly 

 included in the phenomena of conduction; for 

 atom-to-atom radiation is no other than con- 

 duction proper. 



Two methods of measurement may be employed : — 



(1) The time necessary for the production of a given ex- 

 pansion. 



(2) The expansion produced in a given time. 



Subjoined are the numbers derived from a preliminary series 

 of experiments. They show, in millimetres, the depression pro- 

 duced in the tube T when A is maintained for one minute at a 

 given temperature above that of the air. These numbers have 

 accordingly at present no absolute positive value. To have this, 



