2 8 D O C U M E N T 3 J U N E 1 9 2 5 too much, this vessel had to be evacuated. This necessity cost us the most time, because the vessel had to be very strong and completely airtight. Originally, I had intended to generate a good vacuum using an oil pump. Since the oil pump broke down, however, we settled for producing a vacuum with a good water-jet pump (15 to 26 mm Hg). In these circumstances, the insulation is such that within 2 minutes the internal gas tank loses only a small fraction of its charge. If Ohm’s law is satis- fied in the process, this entirely suffices. For this experiment, the tank was charged to 0.25 and 0.5 volt. (After I had left Brussels, I thought that it would have been good to test exactly how far Ohm’s law is satisfied. If the saturation current was already being approached with 0.01 volts, the insulation would have to be checked more closely. But I do believe that we were still far away from the saturation cur- rent in the rarefied gas. Nevertheless, I will still verify this one point.) Now the results: Sensitivity of the electrometer: 1° = 1/40 volts capacity 50 cm volume of the gas tank 27 l pressure difference 7 atm weight of the CO2 = 7·27·2 = 380 g electrometer deflection in the case of a charge of 0.1 stat. units per kg mass would be °. Motions between 0.1 and 0.5° were ob- served, which were always within the order of magnitude of the motion. In the last experiment, the response before and after the experiment was less than ±0.1°/min- ute. The electrometer’s shift during the outflow period was 0.1°. I believe that thus the purpose of the examination has been reached, as any charge is certainly less than 0.01 stat. units/kg. If possible, I am going to perform experiments with a better vacuum and a more sensitive electrometer. If all goes well, one more decimal place may perhaps be gained. The control test with H2 that you suggested was performed and yielded a nega- tive result so one cannot fear any piezoelectric effect. (In the process, I made the unanticipated observation that the inflowing H2 was not electrically charged.) I hope that you will agree with my conclusions so that the only remaining issue to be discussed is how the result can be published. I do not know whether you in- tend to participate in the Swiss scient. convention at the beginning of August in Aarau. This would seem to me to be the most suitable occasion for publication, as the whole problem was broached in Lucerne.[4] Thus I would like to inquire whether you would agree to an announcement in the following form: Einstein, Pic- card, and Kessler.[5] On the Ratio of Static Charges of the Proton and Electron or on the Determination of the V’s. Personally, I would prefer this form for the announcement. I shall, of course, second your own suggestion entirely. If I take the liberty of making a suggestion first, it is merely to save time, as the talks must be announced by 15 June at the lat- est (to Prof. Perrier in Lausanne).[6] 0 1 , 50 -------- - 0 38- , 1 ----------- 300 40 ⋅ ⋅ ⋅ 9 2 , =