[26] See Stallo's excellent Concepts of Modern Physics.
[27] See, for instance, the last new thing in this style—the Helmholtz molecule as improved upon by Sir William Thomson; it is described as follows: "A heavy mass connected by massless springs with a massless enclosing shell; or there may be several shells enclosing each other connected by springs with a dense mass in the centre (far more dense than the ether)." It is not, of course, seriously maintained that this nonsensical creation exists—but that if it did exist it would account for certain unexplained phenomena in the dispersion of light, etc.
Later still (1920) we have the following delightful verdict on the Structure of the Atom, given by Sir Ernest Rutherford—and which I commend to all lovers of clear thinking:—
"The Bakerian Lecture was delivered yesterday before the Royal Society by Sir Ernest Rutherford, whose subject was 'The Nuclear Construction of the Atom.' He said that during recent years much attention had been paid to the nature and structure of atoms. The atomic theory of matter had been definitely proved. The mass of the individual atoms, and the number in any given weight of matter, were now known with considerable accuracy. Not only was matter known to be made up of atoms, but electricity was also atomic in nature, and there was a definite unit of electrical charge which could not further be subdivided. The negative electron, which was a constituent of all atoms of matter, was probably nothing more than an isolated unit of negative electricity, and its small mass was electrical in origin. It had long been considered probable that the atom is an electrical structure, consisting of positive and negative particles, held in equilibrium by electric or magnetic forces. In recent years evidence had accumulated that an atom consists of a positively charged nucleus surrounded at a distance by a distribution of electrons to make it electrically neutral." (From The Morning Post of June 4, 1920.)
[28] The very fact alone that the degrees on a thermometer are equal space divisions shows that they must bear a varying relation to the total volume of liquid as that expands from one end of the tube to the other.
[29] A statement obviously applying—from what has been already said—at only one point in the scale.
[30] I am not, of course, here arguing against the use of thermometers or other instruments for practical purposes. This is certainly the legitimate field of Science. But (as in the case of prediction before mentioned) the exactness of results obtained is a very different matter from the truth of the generalities which are supposed to underlie these results. In using a thermometer you need not even mention the word "temperature."