PATTERX-BLENDIXG. 687 



also brought about by scalloping of the margin and by a fringing 

 of projecting scales ; other methods axe utilised, but these two are 

 specially mentioned as they enter the chosen examples. 



Examination of the wings of insects showed that this marginal 

 pattern is often a short distance from the extreme edge ; expei'i- 

 ments were therefore carried out to discover at what distance 

 the marginal pattern would be effective. A grey square (text- 

 fig. 3, 6) was examined against vaiying backgrounds with the 

 checkered margins (equal in tone after blending) placed at 

 varying distances opposite free margins, as shown in the figure. 

 It was found that the marginal blurring, produced by the 

 checkei'ed patterns, was effective when they were placed a very 

 short distance from the margin. Up to the width of the squares 

 used some efiect was noticeable, but was much more mai'ked 

 when the distance was j^roduced to one half or a quarter of this. 

 The single row of squares appeared to be almost as effective as 

 the double ; attempts were made to obtain more definite results, 

 but in the absence of some method for estimating the amount of 

 blurring, this was found to be impossible. Text-fig. 4, 6, is an 

 example in which the marginal pattern is separated a short 

 distance from the free edge. 



As eye-spots are frequently to be seen near the margins of the 

 wings in Lepidoptera, it was tliought that they might play some 

 part in concealment of form by means of outline-blurring. A 

 few experiments were carried out to discover whether this be 

 so or not. In the centre of a small grey area, an eye-spot was 

 placed consisting of a black centre surrounded by a white ring, 

 the proportion of black to white being so arranged that, after 

 blending, they were equal in tone to the surrounding grey ; a 

 similar gre}'^ area was prepared but with no central eye-spot : 

 these two squares were then examined against various back- 

 grounds ; at the same time the squares were gradually reduced 

 in size : it was then found that when the square (text-fig. 3, 11) 

 was reduced to 12'75 mm., the edges began to be blurred as 

 compared with the edges of text-fig. 3, 12, especially against 

 backgrounds not widely different in tone. Further reduction of 

 the square to 10'3 mm. heightened the difference. 



In this particular experiment the size of the eye-spot was: 

 total diameter 6-75 mm., black centre diameter 4 mm. Eye- 

 spots of different diameters were also used, and it was found that 

 tlie larger the spot, the greater the distance at which mar- 

 ginal blurring was produced. Some experiments were also 

 made, in which the eye- spots were drawn so that after blending 

 they were difierent in tone from the surrounding grey. It was 

 found that they were not very effective in margin -obscuring 

 unless the background was closely similar to their tone, after 

 blending : they appeared as either light or dark areas on the grey 

 square. An experiment was next devised to discover whether 

 this outline-blending, by means of an eye-spot, was more effective 

 than other patterns. As shown in text-fig. 3, i-?, a irrev square 



47* " 



