QQQ SIR DA\ 7 ID BREWSTER ON THE KNOWLEDGE OF 



C, D, are united as in the figure, it is obvious that the flower is incomplete, a 

 part of the central circle of the corolla having been cut off from each half. 

 If we now, by straining the eye, unite C D with B, and also with A, then, at the 

 same time, E will be united with the second or left hand image of A, and G 

 with the second or right hand image of B. But since a piece has been cut out 

 of C D, the half a a of A is nearer the half D D than the other half a a is to the 

 other half C C ; and, in like manner, the half b b of B is nearer the half C C than 

 the other half j3 /3 is to the other half D D. Hence, when the strained eyes unite 

 a to D D, the binocular centre is more remote than when a a is united to C, and 

 the same is true of the other halves ; consequently, the halves D D and b b must 

 appear, as it were, sunk in the wall, or as farther removed from the observer ; 

 and if the defective cutting exists along the line RS from the floor to the 

 ceiling, the whole stripe of paper between R S and P, from the floor to the 

 ceiling, will appear sunk in the papered wall. But if the defect is confined to a 

 portion only of the flowers, then a rectangular space of the breadth R 0, and of 

 a height equal to the defective portion, will appear sunk in the paper. If every 

 junction has the same defect as that at R S, then the whole will appear to 

 consist of equal stripes, every alternate one being raised and the other depressed. 

 In the preceding example, there are only two flowers in a breadth, and 

 their distance is 10^ inches, which is also the breadth of the sunk stripes. But 



21 9 l 

 if the flowers are three or four in number, and their distance -=-, - inches, the 



o 4 



sunk stripes will vary according as we unite two flowers whose distances are 

 in the one case 7 or 14 inches, and 5 or 10^ or 16f or 21 in the other. Calling 

 B the breadth of the paper, n the number of flowers or figures in that breadth, 



T> f) T> O T> 



and W the width of the sunk stripe, then we have W= - or or according 



n n n 



as we unite the two nearest, or the first and second flower, the first and third, 

 or the first and fourth. When W=B, the sunk stripes will cover the whole 

 paper, and all the flowers will lie in the same plane. 



These results afford an accurate method of examining and discovering de- 

 fects in the workmanship of paper-hangers, carpet-makers, painters, and other 

 artists whose profession it is to combine a series of similar patterns in order to 

 form an uniform and ornamental surface. The smallest defect in the similarity 

 and equality in the figures or lines which compose a pattern, and any difference 

 in the distance of the single figures, is instantly detected ; and, what is remark- 

 able, a small inequality of distance in a line perpendicular to the axis of vision, 

 or in one dimension of space, is exhibited in a magnified form as a distance co- 

 incident with the axis of vision, and in an opposite dimension of space ! 



At the commencement of this class of experiments, it is difficult to realize, 

 and very easy to dissolve, the singular binocular picture which we have been 



