connected with the Boiling of Liquids. 91 



another thin layer, which in like manner is also repelled ; and 

 thus the temperature of the liquid rises rapidly by a kind of 

 convection different from that which takes place when the source 

 of heat is limited to the bottom of the vessel. Each thin shell 

 or layer thus repelled rises to the surface, flows over it, and 

 forms a thin vapour-forming layer, not sinking immediately, or 

 rather not being overlaid by the next rising layer of heated fluid 

 until its temperature has been lowered by the giving off of 

 vapour ; it then streams down as if it were colder and heavier : the 

 surface which is nearly on a level with the surface of the liquid 

 in the bath is, from the mode of applying the heat, somewhat 

 lower in temperature than the rest of the liquid, while at 

 such surface the vapour, however much the liquid be super- 

 heated, tends to adjust itself to the true boiling-point of the 

 liquid, although, as will be shown presently, the vapour may 

 readily be superheated. 



As to the highest point to which liquids can be superheated, 

 there are certain statements which seem to me to belong to the 

 region of the marvellous rather than to that of fair inference and 

 common sense. Thus Dufour dropped water into an oil-bath 

 heated to 178° C. (352°"4 F.), and supposed that the water was 

 of the same temperature as the bath because it remained in it 

 without boiling ; and he points with apparent complacency to 

 the fact that at that temperature the elastic force of aqueous 

 vapour is equal to eight or nine atmospheres. De Luc had 

 already performed a similar experiment (the highest temperature 

 of his oil-bath was 257° F.); but with commendable caution he 

 remarks that he was by no means satisfied that the water ever 

 reached that temperature. " Ces gouttes d'eau, renfermees 

 dans Phuile, pouvaient etre dans un etat particulier;" by which 

 he meant " the spheroidal state," although that term had not yet 

 been given to a phenomenon which De Luc understood in its 

 leading features. The doubt expressed by him touches the 

 weak side of most of the statements on superheating. Phy- 

 sicists have supposed that a liquid plunged into a hot bath must 

 necessarily acquire the temperature of that bath. Thus in 

 Donny's celebrated experiment water contained in a tube 8 mil- 

 lims. in diameter, and bent upwards in such a way as was supposed 

 would prevent convection, was boiled in order to expel the air 

 and sealed at the other end, which terminated in two bulbs. 

 The end containing the water was then immersed during three 

 minutes in three baths of calcic chloride in succession, the first at 

 113° C, the second at 121°, the third at 128°; it was then 

 transferred to a fourth bath at 132°, which rose to 138°; but 

 the tube had not been in this bath more than two and a half 

 minutes when the water suddenly burst into steam which was 



