200 O. N. Rood on Stereoscopic Experiments. 



of the thick lines in the other, where the former remained plainly 

 distinguishable. With another drawing, in which the thick 

 lines were six times as broad as the fine, the latter still combined 

 with the edges of the former, but the relief was obtained only 

 with effort, or by excluding most of the light from the slide. 

 With the secondslide a certain amount of lustre is perceived. 



OO 



3 effect is produced by combining in the stereo- 

 such as I and IL the diameter of I correspond- 



scope two circles, such t 

 ing to the inner diameter of II; the smaller circle is seen to lie 

 in the plane of the paper, while the larger circle is considerably 

 inclined to it.* 



The analysis of this singular fact seems to be as follows: 

 when we place in the stereoscope a slide such as that in fig- 2, 

 the difference in the diameter in the circles being yV o^^^ ^° \ 

 combination alternately and rapidly takes place between a and 

 (fig. 2) b, and between a and b', the resultant impression being, 

 according to the principles so ably developed by Prof Wm. i^- 

 Eogers, that of two circles intersecting at a certain angle, the 

 smaller one being situated in the plane of the paper. 



Now, when the drawings are made in fine and thick lines, as 

 in fig. 1, the fine line combines alternately with the exterior and 

 with the inner edge of the thick line, and produces the same 

 efiect. 



2. On the Production of Lustre in the Binocular Combination of sma 

 and large Surfaces. 



By far the most valuable observations on the binocular com- 

 bination of large and small drawings with which I *"^J 

 quainted, are those made by Prof. William B. Kogeij. iD« 

 important distinction between the union of drawings diftering 

 in size in a horizontal or vertical direction, was first pointed o 

 by him, as well as the fact, that 

 while the former case is normal 

 and of daily occurrence, the latter 

 takes place to a much smaller de- 

 gree, if at all. Thus he showed 

 that two such parallelograms a 



\ ^ I — ' 



