practice in observation. Nevertheless, in his own final conclusions, Wirth neglects this practice-effect altogether, and directly compares with one another the results expressed in the various schemata. Besides, I had, for my own satisfaction, constructed individual frequency curves for all the different distributions of attention, and had found great similarity among them; so that for my particular purpose I saw no further necessity of separate treatment. However, even if the different stages of practice had influenced the results, I might justly have followed Wirth's example in his application of the principle of analogy, and might have assumed that longer practice would probably eliminate large irregularities in the frequency distribution.
In conclusion, I still maintain that Wirth has failed to solve his problem, not because, as he tries to show, I had been mistaken in its nature, but because, as I have said before, his numerical clearness values were "obscured or invalidated by complicating factors," and because "the restricting conditions of his experimental arrangement" served only to enhance "the impossibility of overcoming difficulties of observation" whose existence he himself has frequently admitted.
