Of course there were very great individual differences. It is evident that a person who would have remembered all the sixty-four advertisements on this basis of calculation would have made 128 points. The maximum which was actually made was in the case of two women, each of whom reached 50 points. One man reached 49. The lowest limit was touched in the exceptional case of one woman who made only 11 points. The average was 28.4. These figures seem small, considering that less than a fourth were kept in mind, and even by the best memory less than a half, but it must be considered that in the modern style of advertisement the memory is burdened with many side features of the announcement, and that the result is therefore smaller than if name and article had been memorized in an isolated form. But these figures have no relation to our real problem. We wanted to compare the memory fate of the advertisements on the one kind of pages with that of the parallel advertisements on the other kind. As soon as we separate the two kinds of reproduced material we find as total result that the forty-seven persons summed up 570 points for the advertisements on pages with comic pictures, but 771 for the advertisements on pages which contained nothing else. The average individual thus remembered about six whole advertisements out of the thirty-two on the combined pages, and about eight and a fifth of the thirty-two on the straight pages. Among the forty-seven persons, there were thirty-six who remembered the straight-page notices distinctly better than the mixed-page advertisements, and only eleven of the forty-seven showed a slight advantage in favour of the mixed pages. In the case of the men this difference is distinctly greater than in the case of the women. Only two of the fifteen men who participated showed better reproducing power for the mixed material, while nine of the thirty-two women favoured it. As the advertiser is not interested in the chance variations and exceptional cases among the reading public, but naturally must rely on the averages, the results show clearly that the propaganda made on pages which do not contain anything but advertisements has more than a third greater chances, as the relation was that of 6 to 8.2.