Astronomical and Nautical Collections. 44& 



veral yards : and this new hypothesis would introduce a 

 multitude of difficulties still more embarrassing than the one 

 in question. 



In all such experiments, the margin of the shadow is 

 80 confounded with the fringe of the first order, that it is 

 impossible to determine its limits by the eye only. I em- 

 ploy, for making the shadow, a metallic wire or cylinder, 

 large enough to prevent any confusion from the effect of 

 the light inflected at the opposite side ; a point which I 

 ascertain by enlarging its breadth by the addition of a small 

 substance at one part, and finding that the external fringe 

 opposite to it remains unaltered. Then, if I wish to ascer- 

 tain, for example, the situation of the darkest part of the 

 third order of fringes, with respect to the geometrical 

 shadow, 1 measure the distance between the two dark 

 stripes of the third order on the opposite sides of the 

 shadow, and subtracting from this the computed breadth 

 of the shadow, one half of the remainder will obviously be 

 the interval required. The diameter of the wire is mea- 

 sured by a gage, or beam compass, which allows me to 

 read off the thousandths of an inch by its vernier, and to 

 estimate less than half of this space with sufficient accuracy : 

 and the instrument itself has sometimes been used instead 

 of a wire, taking care that the two plates, or cocks, should 

 be far enough separated to avoid any confusion from their 

 mutual effects, and then deducting the interval of the 

 opposite fringes from the breadth of the geometrical shadow, 

 in order to obtain the double distance required. 



The curvilinear motion of the fringes can only be satis- 

 factorily explained by the mutual influence of the luminous 

 rays, whatever theory we may adopt respecting their nature: 

 from this principle only can we conceive how it happens, 

 that the rays inflected or diffracted, in the neighbourhood of 

 the body, can, without ceasing to be propagated in right 

 lines, give rise to curvilinear trajectories of dark and bright 

 stripes: and it is sufficient for this purpose that the different 

 points, in which they are the most strengthened or weakened 

 by their combination, should be situated in curved lines, 

 and not in straight lines. This would happen, for example. 



