620 



SCIENCE. 



[N. S. Vol. XII. No. 304. 



3. Since the distances x and y are large 

 (2-10 meters) , it will be possible to obtain 

 gratings of different fineness (effective hori- 

 zontal distance of wires apart) by merely 

 rotating either grating on an axis parallel 

 to the wires. Since the focal planes have 

 now been shown to be real, it is expedient 

 to project the whole phenomenon with sun- 

 light, and if parallel rays are not wanted a 

 ground glass screen or better, a screen of 

 scratched mica which is more translucent, 

 may be interposed at Cin Fig. 1, in front 

 of the first grating, A. Thus if L be the 

 direction of sunlight and the angle of 

 rotation of either grating, the figure meets 

 the present case. If A be left normal and 

 B rotated, results are obtained for the case 

 where the projecting meshes are smaller 

 horizontally than those projected. If jB be 

 left normal and A rotated, the projected 

 meshes are the smaller. For any angle 6 

 of either A or B, the grating B and screen 

 S may be moved along the axis to locate 

 the other focal planes for the same mesh 

 ratio. With the proper angle 6 images may 

 be focused for any distance y relative to x. 



TABLE 3. — DATA FOK A FINEE PROJECTING MESH 



(B rotated), a; = 200 cm. a = l. 



At long ranges (500 cm. and more) the 

 white shows faint interference fringes usu- 

 ally with a pink center. At 7 meters, when 

 the ground glass screen is interposed in 

 front of the first grating, A, the effect is 



a remarkably clear diffraction pattern fully 

 two feet square or more, consisting of nar- 

 row, strong, black lines on a dull white 

 ground. When the grating space of B is 

 reduced to ^ by rotating it, very fine lines 

 fainter but very clear show on the same 

 ground. For other mesh-ratios the field is 

 blank, and sharp adjustment of is neces- 

 sary. Diffuse, non-parallel light, therefore, 

 is equally active, and being free from the in- 

 tense but circumscribed glare of full sun- 

 light, gives more striking results. Moreover, 

 the same figures as above show through 

 the dull mica screen for all the distances 

 noted in the table. 



Special attention may be called to the 

 fact that the figure is still distinct even at 

 a distance of 30 meters between the image 

 S and the projecting grating B. 



The results of the following table were 

 obtained by keeping grating B normal and 

 rotating A. 



TABLE 4. — data FOR A COARSER PROJECTING 

 MESH {A rotated), a; := 200. 6^1. 



As the obliquity of A is increased the 

 focal plane frequently does not sharply van- 

 ish, the image merely becoming smaller. 

 Because of this indefiniteness of smaller 

 images further measurement was not at- 

 tempted. It will be seen that the angles 6 

 for the same y do not correspond to the 

 preceding table, as was directly proved by 

 exchanging the gratings. This is the im- 

 portant datum of the new series of obser- 

 vations, and makes it needless to adduce a 

 greater number. 



SCHEME FOE THE PROJECTION OF ONE GRAT- 

 ING BY ANOTHER. 



4. In order to interpret these results it 

 will be expedient to introduce a simple 



