§ 1. Mechanical Analogy.

of a gas, which perceives internal friction and heat as phenomena that only appear significantly different in stationary and approximately stationary states. However, in certain transitional cases (very rapid sound vibrations with heat development, friction, or heat conduction in very dilute gases¹) a sharp distinction is no longer possible between what is visible motion and what is heat motion (cf. § 24). This is exactly like in Maxwell’s theory of electricity, where in transitional cases the separation of electrostatic and electrodynamic forces, etc., can no longer be maintained. It is precisely in these transitional areas that Maxwell’s theory of electricity produced something entirely new; likewise, in these transitional cases, the gas theory leads to completely new laws, which make the ordinary hydrodynamic equations—corrected for friction and heat conduction—appear as mere approximation formulas (cf. § 23). These completely new laws were pointed out for the first time in Maxwell’s treatise published 16 years ago, "On stresses in rarefied Gases." Among the phenomena to which a theory limiting itself to the description of old hydrodynamic appearances could never lead, the radiometer effects The movement of vanes in a vacuum caused by light or heat, discovered by William Crookes in 1873. are especially to be included. Attempts to observe them under entirely different conditions and quantitatively would surely provide proof that the inspiration and guidance in a certain, previously neglected area of experimental research can only originate from the gas theory; after all, the enormous fertility of Maxwell’s theory of electricity for experimental research also remained almost unnoticed for more than 20 years.

While in the following, any qualitative difference between heat and mechanical energy is excluded, the old distinction between potential Stored energy based on position and kinetic energy Energy of motion shall be maintained when considering the collisions of molecules. However, this distinction does not at all get to the essence of the matter. The assumptions regarding the interaction of molecules during a collision have entirely the character of something provisional and will surely

¹) Cf. Kundt and Warburg, Poggendorff's Annalen A major German scientific journal of the 19th century 155. p. 341. 1875.