Heat across Layers of Gas. 441 



which is familiar to us as the way in which a laundress tests 

 the heat of her smoothing-iron, and which was studied by M. 

 Boutigny under the name of the spheroidal state of liquids. 

 If a drop of water or other cold volatile liquid is allowed to 

 fall into a smooth and sufficiently hot metal dish, it continues 

 liquid instead of flashing off into vapour, and exhibits an ap- 

 pearance of great mobility. Here the liquid settles down upon 

 the Crookes^s layer which envelops the metal, and reduces 

 that portion which is under it to the condition of a compressed 

 Crookes's layer. Now the mechanical peculiarity of a com- 

 pressed Crookes's layer is, that it exerts more force in the 

 direction along which the heat travels (in the present instance 

 up and down) than in the perpendicular direction ; and 

 inasmuch as the pressure sideways must continue to be the 

 pressure of the atmosphere, the excess of pressure upwards 

 is able to support a weight. When we also remember that 

 this excess of pressure would be augmented by still further 

 curtailing the Crookes's layer, ^. e, by depressing any part of 

 the drop, we have all the mechanical conditions necessary for 

 the stable equilibrium of the drop, if only the force rises to a 

 sufficient amount before the drop settles down quite through 

 the Crookes's layer. This, by the theory, depends altogether 

 on the difference of temperatures which can be maintained. 

 The first thermal effect is, that the drop becomes warmed by 

 the radiation and penetration of heat from the hot metal below. 

 This causes the liquid, if volatile, to lose heat by evaporation, 

 and, in most cases, to lose a little heat also by radiation to 

 surrounding bodies. As the temperature of the drop rises, 

 the heat thus lost increases, while at the same time the heat 

 received from below diminishes, and if a balance between the 

 two is effected before the liquid reaches the boiling-point, the 

 drop continues liquid, the temperature remains henceforth 

 unchanged, and we have before us the striking spectacle of a 

 liquid in the spheroidal state. 



21. I have long thought it likely that the drops which may 

 be sometimes seen running over the surface of a volatile liquid 

 are supported by compressed Crookes's layers intervening be- 

 tween them and the liquid on which they float — that they are, 

 in fact, drops in the spheroidal state ; and the following ob- 

 servations abundantly confirm this suspicion. These floating 

 globules are easily formed when a liquid as volatile as spirits 

 of wine is allowed to fall in drops of a medium size from a 

 height of about 8 centimetres into a vessel containing some of 

 the same liquid moderately warmed*. They can also be 



* . A better way to form tliem is gently to lay the drop as it were with 

 a fine pipette upon the surface of the heated liquid. The di'op when 



