@article{oai:nagoya.repo.nii.ac.jp:00008320,
 author = {後藤, 倬男 and GOTO, Takuo},
 journal = {名古屋大学文学部研究論集　哲学},
 month = {Mar},
 note = {When two center circles (CCs) of identical size are presented side by side either in Ebbinghaus (E) illusion or in Delboeuf (D) illusion (Fig.1), the sizes of these CCs look different as the size contrast or assimilation interacts either between the surrounding circles (SCs) and the CC in E illusion or between the concentric outer circle (COC) and the CC in D illusion. This study examines the underlying relationship between E and D illusions, which have a similar appearance, by presenting mixed stimuli of E and D illusions, and the magnitudes of illusion (MIs) were obtained by two kinds of measurement. The black stimulus figures (E and D illusions) were printed on white sheets of paper (211×298mm) and presented in a form of "booklet" to the subjects. The MIs of the CCs (dia: 20mm) in E and D illusions were measured in one session using two methods; production and magnitude estimation. These illusions were observed by the subject located 40 cm apart. The size ratios of E illusion (REs) and D illusion (RDs) are as follows. REs: between the SCs and the CC-1/4, 1/2, 1/1, 3/2, and 2/1; RDs: between the COC and the CC - 4/2, 5/2, 7/2, 9/2, and 11/2. These ten stimulus figures with different size ratios from two illusions (E and D) were presented randomly in one session with the control figure (single circle) presented at the beginning and the end of each session. Subjects, 20 unversity students in the introductory psychological course, took part in this experiment. In one session, they were asked to observe 12 figures within ten minutes using two measuring methods and they performed seven sessions consecutively over seven weeks, once every week. This experiment has revealed the follwign three results. (1) The MIs of E and D illusions showed no significant difference when they were measured using two kinds measurement (production and magnitude estimation). This means that these two methods are useful in measuring the MIs for many untrained perticipants with use of the booklet method. (2)With E illusion, the MIs displayed characteristic variations as a function of the RE between the SCs and the CC (Fig.2). These variations reflected clear size contrast between the SCs and the CC as we suggested in our previous studies (Goto, 1978, 1980a, b, 1982, 1987, 1990, 1991; Goto & Ohya, 1989). (3)The MIs of D illusion shifted from overestimation to underestimation as the RD between the COC and the CC increased (Fig.3). This tendency was almost the same as those found by Morinaga (1935) and Ogasawara (1952). However, the transition (zero point) from overestimation to underestimation occurred when the RD was quite smaller than the RDs obtained in their experiments (Fig.4). In conclusion, this experiment using the two kinds of measurement with the booklet method has again verified that E illusion is size-contrast oriented. The traits of MIs variations in E and D illusions looked alike each other if the COC of D illusion was matched with the "estimated" outer circle in the SCs of E illusion. Consequently, these two illusions seem to originate from the common factor as Coren & Girgus (1974, 1978) already indicated. Although they included E illusion into D illusion to explain the cohesive contour interaction between the COC (SCs) and the CC, I would argue that D illusion should be included into E illusion because the size contrast is more dominant to cause mainly these illusions. Accordingly, this D illusion is caused due to the size contrast between the COC and the CC, and the cohesive factor may partly work only when the COC (SC) and the CC are positioned close each other and look like a ring (Ogasawara, 1952). However, this study suggests that the other factor of such as "apparent depth" between the SCs and the CC may play an important role as the supporting factor throughout these size illusions.},
 pages = {101--112},
 title = {大きさの円対比錯視（Ebbinghaus 錯視）に関する実験的研究（Ⅴ） : Ebbinghaus錯視とDelboeuf錯視の関係についての一考察},
 volume = {40},
 year = {1994}
}