Graham was not as far as we can learn endowed, like Davy, with the sensitive nature of a poet, yet his work on hydrogen proves him to have possessed a large share of the gift of imagination. Picturing to himself the hydrogen atom as essentially similar in its chemical functions to the atom of a metal, he tracked this light invisible gas through many tortuous courses: he showed how it is absorbed and retained (_occluded_ as he said) by many metals; he found it in meteors which had come from far-away regions of space; and at last, the year before he died he prepared an alloy of palladium and the metal hydrogen, from which a few medals were struck, bearing the legend "Palladium-Hydrogenium 1869." Within the last few years hydrogen has been liquified and, it is said, solidified. Solid hydrogen is described as a steel-grey substance which fell upon the table with a sound like the ring of a metal. But Graham's most important work was concerned with the motion of the ultimate particles of bodies. He uses the word "atom" pretty much as Dalton did. He does not make a distinction between the atom of an element and the atom of a compound, but apparently uses the term as a convenient one to express the smallest undivided particle of any chemical substance which exhibits the properties of that substance. As Graham was chiefly concerned with the physical properties of chemical substances, or with those properties which are studied alike by chemistry and physics, the distinction between atom and molecule, so all-important in pure chemistry, might be, and to a great extent was, overlooked by him. In considering his work we shall however do well to use the terms "atom" and "molecule" in the sense in which they are now always used in chemistry, a sense which has been already discussed (see pp. 139-143). Many years before Graham began his work a curious fact had been recorded but not explained. In 1823 Doebereiner filled a glass jar with hydrogen and allowed the jar to stand over water: on returning after twelve hours he found that the water had risen about an inch and a half into the jar. Close examination of the jar showed the presence of a small crack in the glass. Many jars, tubes and flasks, all with small cracks in the glass, were filled with hydrogen and allowed to stand over water; in every case the water rose in the vessel. No rise of the water was however noticeable if the vessels were filled with ordinary air, nitrogen or oxygen.