摘要
典型的錐細胞直徑0.5微米，占0.5分視角，也就是2.0的視力。但是對於兩線段是否完全對齊的vernier視標的判斷力，則常常超過這一解剖學上的極限，我們稱之為“超視力”。大部分學者將此現象歸因於網膜內外兩叢狀層神經的側向交互作用。我們以IBM 5540微電腦設計修改vernier視標，以“邊緣”代替“線段”，而心理物理測試的方法，則以強迫選擇，要求受測者決定左側或右側未對齊，或兩側均對齊，或均未對齊。我們先在同樣螢幕上測得藍氏C視標的視力值，發現和傳統C視標測得的視力值吻合。其次改變vernier視標“邊緣”的長度，測量各種邊緣長度下的vernier視力。將vernier視力相對於邊緣長度作圖，我們發現邊緣由長而短，在相當大的範圍內，vernier視力不變。直到邊緣長度小於1.0C視標標缺口五倍左右，vernier視力才開始下降。如果1.0視標代表兩個錐細胞的大小，則我們推論造成超視力的有效神經側向交宜作用，應不超過半徑10個錐細胞的範圍。這和解剖學上，網膜水平細胞和無突細胞的樹狀突廣大的伸展範圍並不符合。
關鍵字
無資料並列摘要
Typical cone size of 2.5 microns subtends 1/2 min-are, which corresponds to 2.0 visual acuity. The judgement of alignment or misalignment of two line segments gets vernier acuity of less than this limit. We call this 'hyperacuity'. Hyperacuity was explained by most scientists as lateral interaction at the retinal outer and inner plexiform layers. We designed modified vernier targets on IBM 5540 monitor. The target showed edges rather as line segments. 4 forced choice answers were misalignment on right side, left side, both sides or neither. We then shortened the edge to decrease the lateral effect and measure vernier acuity under different edge length. Plotting the vernier acuity vs. edge length, we got plateau over wide range of edge length. It was not until the edge length got shorter than about five times the opening of 1.0 Landolt C target when the vernier acuity started to decrease. If the opening of 1.0 Landolt target corresponed to the size of two cones, we concluded that the radius of effective lateral interaction subtended no more than 10 cone size, though anatomically, the dendrites of horizontal and amacrine cells extended much farther than this area.