• (With 5 figures in the text)


‘Relative frequency’ recorded by point quadrats measures not the actual area of foliage but the area projected in the direction in which the quadrat lies. Accordingly the relative frequency varies both with the slope of the foliage and also—when inclined quadrats are used—with the inclination of the quadrat. A theoretical study reveals that variation in relative frequency resulting from differences in foliage angle is greatest for vertical quadrats, is considerably reduced when (as suggested by Tinney, Aamodt and Ahlgren) quadrats are inclined at 45°, and is least when quadrat inclination is 32.5°. Accordingly the usual, vertical position for point quadrats is the worst possible one, since it results in the most erroneous estimates of percentage contribution (area basis); while with quadrats inclined at 32.5° errors are greatly reduced and are of an order acceptable in general survey work.