We report basis experimental data of readability about three classes of logical symbols in mathematics OR. .AND. in Fortran and in C (java) first. as the cognitive psychology-measurements method, we propose methods of three classes. these are(1) monadic operation, symbols count experiment, (2) binary relation experiment, (3) Polynominal (N-ary) relation experiment. these experiment's results were observed following the result of these experiments was observed following power law of practice as the acquisition model of skill ( is execution time, p is total practice time, initial skill degree of proficiency b, skilled value y ) [14] [16] [15] in symbols count experiment result, it was 1.6:1.5:1 in the ratio of initial skill, and it was 2.2:2:1 in the ratio of mathematic, Fortran and C at skilled value. and in mathematics was 90% of ratio of a correct answer, others being 95% or more of rates of a correct answer. therefore is easy to misidentified the symbols count is a bad symbols with low rate of a correct answer and cognitive time slow twice as compared with this paper also describes readability of arithmetic sign, predicate logic symbol and negative symbol. the quantity-judging about readability as shown here should be added in the design of writing system of programming language, symbolic and algebraic computation.