In the preceding numbers of this JOURNAL, the writer together with one of his students set forth the results of an investigation on the apparent distance of the negative after-image in the field of the closed eyes. Up to the time of printing, the writers had found no material bearing on the topic save casual references by Fechner and Hering. The fated article which, of course, had a Teutonic existence somewhere, has turned up in v. Graefe's Archiv for 1885 with Dr. G. Mayerhausen as sponsor and as title Ueber die Gr6ssenverhdiltnisse der Nachbilder bei geschlossenen Lidern (Abteilung 2 S. 23-24). The results differ so widely from those published in the JOURNAL as to call either for explanation or discussion or both. Experimenting on himself, Mayerhausen finds that the negative afterimage in the darkened field appears as large as the inducing object when the latter is set at a distance of two meters. In the case of the Stanford observers, three in number, unacquainted with each other's findings, the region of equality lay inside of eighteen centimeters from the cyclopean eye. Mayerhausen's method was as follows: As inducing objects he used four white discs, 1, 2, 4, and 8 cm. in diameter respectively, and developed after-images from them at ten different distances from the eye, beginning with five meters and running down to io cm. When the after-image was at its maximum clearness, he opened his eyes and compared its diameter with the readings of a centimeter scale held in his hand. In this way, he made eighty readings, viz., two for each size after-image at each of the ten distances. The writer humbly submits that this is a less trustworthy procedure than that described in the JOURNAL where the after-image with closed eyes was compared with the image projected on an adjustable screen until a region was found where the two images showed no appreciable difference. Within this region, after-images were induced and their size measured by projecting them on a ruled screen. The equating of the two forms of after-image was almost as accurate as if the images had been actual cardboard disks, successively compared. The estimation of the projected image was harder but nevertheless the size of the mean variation of the readings (from 4% to 10%) for the three observers showed that the judgments were based on actual if rough measurement.