When an oxide‐coated cathode tube of the gas or vapor type passes more than a certain amount of current a spark or cathode spot occurs on the cathode and the tube is damaged. A peak current rating is, therefore, essential for any tube of that type and it becomes necessary to take data on how the sparking current is influenced by cathode temperature, vapor or gas pressure and other factors. By ordinary test methods each spark is so destructive that only one reading can be taken on one tube and the effect of the above factors is confused by the variation in quality of tubes used to obtain the data. The authors have developed a test procedure that so limits the duration of the spark that very little damage to the cathode results. This makes it possible to obtain a large number of readings on each tube and thereby observe effects otherwise obscured. The experimental procedure appears to be a good one for investigations in this field and the results obtained shed some light on the mechanism of sparking.