simple straight line, their motion would be uniform and undifferenced; but, as neither the position nor the size of a mass undergoes change when temperature does not vary, atomic paths must suffer oft-repeated stops. Elastic particles in this state must have incessantly fluctuating velocities, yet always oscillating about a fixed mean. Matter endued with heat cannot have its particles in absolute contact, or the compressibility or contractibility which is the inseparable property of any mass would not exist. For argument's sake, however, let it be admitted that from absolute contact or any other cause heat-motion is a uniform one. If it be a purely axial rotation, then the equators of the atoms move faster than the poles, and the movement is not homogeneous. Exactly so, if the atom describe as an orbit a circle, ellipse, or other figure recurrently. Such motion would involve axial rotation, the atom would resemble our earth, and different parts of it would move with different velocities. In the case of two tangible spheres of like dimensions it is easy to show that, when swiftly moving at an equal rate, the speed of the one can be accelerated at the expense of the other, by applying it at a point not equatorial to the equator of its neighbor. In some such way it is conceivable that differences in molecular motion may widen from those subsisting between the parts of an individual molecule.
The imperfect homogeneity of thermal motion, which is here contended for, has some palpable parallels in the distribution of two other phases of energy—electricity of high tension and magnetism; these forces are cumulative in their manifestations, increasing in intensity toward the poles of the masses presenting them.
Thirdly, it is not strictly an accurate premise in the theory that, when heat is produced from any other force, it is unaccompanied by any phase of energy not thermal. Increments of heat invariably alter the dimensions of bodies, as a rule expand them, and thus part of the original energy applied appears as gravity. The sun in warming the earth's atmosphere lifts it, and, when the air cools, its fall is of no insignificant dynamic value. What is so evident in this extreme case is true of any mass whatever when heated. Not only is heat pitted against gravity, but at times against cohesive and crystalline forces, which, though overcome, must modify and diminish its effects.
There is a check to the continuous increase of temperature which is of much more importance than those just noted, but akin to them. A compound substance receives additions of heat with tolerable evenness up to a certain point, when it is resolved into two or more simple constituents, according to its complexity. These if compound are in turn decomposed into their chemical elements if more heat be applied. Now, chemical energy is a motion quite distinct by itself, and we find that heat in its higher degrees must coexist with it. So that on this account we cannot accept the notion that heat is ever to become the only kind of motion in the universe. In so doing we recog-
