523of reading, I n the first experiment we observed for some assistants such an effect, which we interpreted as fatigue. Now none of the assistants was working under pressure, since only half the amount of work had to t)e done. Bc) effect of order of session was observed, It t has already been mentioned that there was a si~;nitic;int reading-to-reading variation for assistants G, L, and 0. Session-to-session variation was signiticant for two assistants, Q and S, and nearly significant for L and ~l. Day-today variation was significant for two assistants, I' and R, and relatively large values were estimated for (~, L, and T (see Table II).There seem to be only two assistants who show no effect of any factor, namely II and K. For three (1, J, X) there are indications only of some day-to-day variation: for all the others some vari-8.