IN a letter in NATURE of September 30, Mr. Arvon Glynne describes an elegant and simple method for the comparison of four-terminal resistors. In it, as he points out, two of the ‘potential’ leads have to carry a small current. I would like to direct attention to another method, almost as economical in apparatus, in which no current passes in any of these leads. Resistors of about 1/1000 ohm can be compared in this way with an accuracy of one part per thousand, using apparatus ordinarily available in a laboratory. The principle will be evident from the accompanying diagram.