418 Prof. Magnus on the Influence of the Adhesion of 



the tubes become coated with water in their entire length when 

 moist air is forced in. When a line thermopile was placed on 

 the outside of one of the metal tubes used, in such a manner that 

 the line of contact was parallel with the axis of the tube, it be- 

 came heated whenever moist air was forced in, while it indicated 

 cooling when dry air entered. This heating did not take place 

 only when the pile touched the outside of the tube opposite the 

 aperture through which air entered the tube, but also when it was 

 placed in any other position, provided it actually touched the tube. 



Considering how much more heat is absorbed by water than 

 by a polished metallic surface, it is clear that the inside of the 

 tube absorbs a far greater proportion of the rays which are inci- 

 dent upon it when it is coated with water than when it is quite 

 dry. The more heat the tube absorbs, the less does it reflect, 

 and the less reaches the pile. The action of water in these tubes 

 is probably greater owing to its not forming a continuous layer, 

 but being deposited in individual small masses, which scatter in 

 all directions the small quantity of heat which was not absorbed. 



The quantity of heat which reached the pile, owing exclusively 

 to reflexion from the inside of the polished brass tube which was 

 filled with dry air, was about six times as much as that which 

 fell upon it when the tube was omitted. But if six-sevenths of 

 the heat which the pile receives be reflected, any diminution of 

 reflexion must necessitate just as great a diminution of heating. 



The particles of carbon which coat the inside absorb like the 

 condensed particles of water, but in a higher degree ; hence its 

 absorption is increased but little, or not at all, by the deposited 

 water. For this reason no cooling is observed when moist air is 

 forced into the blackened or the velvet-coated tube ; a heating, 

 on the contrary, was observed. This is manifestly due to the 

 condensation of vapours on the inside of the tube. For when no 

 source of heat was used, and moist air was blown into one of 

 these tubes, the pile was also heated. With polished metal 

 tubes this heating effect could seldom be perceived, doubtless 

 because the heat as soon as it became free was conducted through- 

 out the entire mass of the metal. With badly conducting lamp- 

 black this could not occur, and still less with velvet ; hence by 

 using the velvet tube the heating was even stronger than with 

 the lampblack. 



This heating did not arise from a motion of the moist air to 

 the pile ; otherwise it must have been observed in all these 

 tubes, for they were all of the same dimensions. Moreover it 

 must not be omitted that the heating which the moist air pro- 

 duced in these tubes when the source of heat was wanting, was 

 greater than the increase of heating which ensued by forcing 

 in air when the heated cube was used. 



