PATTERX- BLENDING. 685 



different toned backgrounds, and it was found that on all occa- 

 sions marginal indistinctness after pattern-blending occurred 

 where the black and white areas were interrupted at the edges ; 

 for instance, fig. 2 appeared as a grey square whilst fig. 5 showed 

 an ill-detined grey area fading into the background. It was 

 noted that when tlie background was near in tone to the squares 

 after blending of the pattern, this obscuring effect was most 

 marked : when the background was far removed in tone, for 

 instance was black, or white, very little obscuring was noticeable. 

 Other experiments showed that the greater the width of the 

 black and white areas passing out at the margin, the greater the 

 obscuring effect after pattern-blending. 



One is now in a position to conclude that the method of 

 obliterative shading by pattern-blending, as seen for instance in 

 the Zebra, if it does not possess an advantage over the usual 

 method as regards visibility, as measured by the distance at which 

 the object is visible, nevertheless would seem to have a decided 

 advantage in that the contour or silhouette of the animal against 

 backgrounds approaching it in similarity, is made blurred rather 

 than sharp. It is noteworthy that the stripes of the Zebra 

 everywhere pass out at the margins (see text-fig. "l^ 1, 3, and 7). 



Referring to the possibility of the Zebra's stripes being 

 pictures of reeds in high light and shadow, as has been suggested 

 by Thayer, it is noteworthy that the black stripes decrease in 

 breadth from above, down, which is contrary to expectation and 

 to what actually is found among other animals. Thayer has 

 pointed out that the markings on the backs of animals are 

 usuall}^ smaller than those on the ventral aspect, because they 

 represent objects in the landscape more distant : they depict the 

 foreshortening of the ground. If reeds are painted on the zebra's 

 back, one would at lenst have expected the stripes to be of equal 

 width. 



The Tiger presents an excellent example of reed-painting : on a 

 coat showing Well-marked obliterative shading, dark stripes are 

 to be seen, they are irregularly distributed, they for the most 

 part increase in width from above, down, and on the belly 

 they terminate in a large dark mass ^vhich, curiously enough, 

 often shows a small white centre (compare text-fig. 2, 1 with text- 

 fig. 2, 6). 



Examples of outline-masking by means of pattei'n -blending at 

 the margins are very commonly seen amongst Lepidoptera, but 

 liefore considering instances, some experiments must be referred 

 to. If one takes two flat surfaces, one an even dark grey and 

 the other an even light grey, and brings them in apposition as 

 shown in text^fig. 3, 6^ the junction will appear sharp, no matter 

 how viewed. If now, at the junction a na.rrow band of black 

 and white squares, which after blending will have the same tone 

 as the light square, be interposed, see text-fig. 3, 6, it will be 

 found that the junction no longer remains sharp ; if the outer 

 square be replaced by areas of different tones varying from light 



Paoc. ZooL. Soc— 1915, No. XLYII. 47 



