52 EINSTEIN'S "MASCHINCHEN" of the glass disk entered A and was earthed by a little chain, it was charged by induc- tion. Then the same sector entered the second (uncharged) box B and was connected to it part of the charge was then transferred to B. The next sector went through the same process. The cycle was repeated until box B had accumulated a large charge. The connections of the machine were then inverted and the charge on B was used to increase the charge on A. In this way, the boxes alternately served as collectors and inductors. Although the machine was able to generate strong sparks, it was not a very prac- tical device. Even though Belli later introduced a number of improvements, his gen- erators did not become very popular and only few of them were constructed in Italy. In 1867 William Thomson proposed a similar instrument, calling it a "replenisher." It was applied as part of his "siphon recorder" and later, with a better design, as part of his absolute electrometer. In 1885, the German physicists Julius Elster and Hans Geitel invented a doubler which was identical to the one described by Belli. Be- cause of its simple design and operation they suggested using it as a didactic instru- ment for illustrating the operating cycle of more complicated induction machines. From 1860 on several other types of electrostatic induction machines were invent- ed. At the beginning of the twentieth century, such machines were continuing to hold their own in the laboratory despite the growing popularity of the powerful electro- magnetic induction coil. The electrostatic machines produced the newly discovered X-rays and electric effluvia generating ozone, as well as contributing to the develop- ment of medical electrotherapy. No longer were the machines used, however, for the purpose of amplyfying charges so that they could be measured. II In a paper published in April 1908 Einstein outlined his method of measuring small quantities of electricity. Although he did not mention any of his predecessors, the ap- paratus he proposed worked exactly like Belli's "macchina ad attuazione." In a dia- gram he sketched a schematic arrangement with two fixed metallic plates A1 and A'1 and a series of moving plates B. The weak charge to be multiplied was on A1. One of the plates B, close to A1, was first earthed and charged by electrostatic induction. B was then moved closer to A'1 and electrically connected to it. Part of the charge on B was thus transferred to this plate. Then the cycle was repeated. By an appropriate choice of the capacity of the systems A1-B and A'1-B it was possible to increase the voltage on A'1 until it reached an equilibrium value. In one respect Einstein's machine (a detector) was different from Belli's instru- ment (a generator): in the latter case the box with the original charge acted first as an inductor and, in the next stage, as a collector where the charges were multiplied. In Einstein's machine the charge of the inductor A1 was not increased but remained the same. His machine was thus an addition machine and not a true multiplier. Ein- stein also proposed to link a series of such devices together in a multiple-stage ma- chine to further increase the multiplication factor and hence the output voltage.