208 ANNUAL OF SCIENTIFIC DISCOVERY. 



Mariotte's law, cohesive action becomes appreciable, and this is proof 

 enough that in masses many layers contribute their action in making up 

 the total cohesion. If we conceive any fluid mass to be distributed into 

 layers, then the correct measure of fluid cohesion will be the force 

 requisite to produce a direct simultaneous separation of all the parts 

 along a unit of the dividing surface between two layers. This is equal 

 to the resultant of all the forces acting from either direction against 

 this unit of surface, these forces being held in equilibric by the equal 

 opposing forces. To present the grounds which seem to me to author- 

 ize the conception that repulsion in all states of aggregation is only 

 exercised between adjacent molecules, while the attractive actions arc 

 the resultant of all the primary constitutional forces, and extend 

 through longer spheres, would involve the exposition of a complete 

 theory of molecular mechanics. I must therefore leave as an assump- 

 tion the conception that in fluids, the only repulsion to be taken into 

 account is the continuous layers, which prevents their yielding farther 

 to the cohesive forces pressing them together. 



I come now to an important deduction from the preceding discus- 

 sion. Fluid surfaces are in a state of weak cohesion, as compared with 

 fluid interiors ; hence, a partially atmospheric condition of rarefaction 

 exists along such bounding surfaces. If, then, we assimilate heat to a 

 molecular repulsion, as is customary, we see at once that as the tem- 

 perature is raised, the weak cohesion in the surface will be overcome, 

 long before the mass is heated to that point which will overmaster its 

 internal cohesion. Hence, the surface molecules will freely pass off as 

 vapors, while a strong cohesion still exists throughout the entire mass. 

 Evaporation thus goes on at surface at all temperatures above that 

 which just suffices to overcome the weak surface cohesion. This con- 

 stitution, or structure, necessarily characterizing the limiting layers of 

 fluids, is the true and full explanation of evaporation in all its forms. 

 From this we see that a fluid mass, without interior or exterior sur- 

 faces, or so enclosed as virtually to answer this description, might be 

 heated up far above the boiling-point without boiling. We see that 

 ebullition is but the effect of an internal evaporation, starting in mi- 

 nute air-bubbles, and growing with the expanding bubble. We see 

 that water, entirely freed from air-bubbles, and with a restricted open 

 surface, as in Donny's tube experiments, should go on heating up far 

 above the boiling point, until at last the whole heated mass Avould 

 flash into steam with an explosion. All the phenomena described by 

 Donny, in his excellent paper in the Annales de Chimie et de Phyxique, 

 follow as easy and obvious deductions from this constitution of the 

 fluid surface. Indeed, we do not wonder at his being forced, from 

 his experiments, to conclude empirically that there must be some pe- 

 culiar quality in surfaces which makes evaporation take place so much 

 more readily on them than on fluid masses. We see, too, how utterly 

 fallacious are the experiments usually taken as measuring fluid cohe- 

 sion in surface layers, which, with the free mobility of fluid parts, fully 

 explains all the observed results. This fully explains how a too per- 

 fect boiling of the mercury in barometer tubes makes it adhere at the 



