26 Prof. Snell on new articles of Philosophical Apparatus. 
also is every time reversed, since the cusp is necessarily turned 
away from the surface which produces the last reflection. It 
should be remarked, that all the orders are never in full view at 
once, as represented in the figure. When the first is most dis- 
tinctly formed, no others are to be seen; and generally, when the 
caustic of either order is brightest, the higher orders are not 
formed at all, and the lower ones only in part. The caustics of 
the first order extend only half round the mirror, and are termin- 
ated by it in opposite points ; but in all the others the two branches 
intersect, and may be traced round much more than the entire 
circumference. In proper positions of the instrument, the branches 
of many successive figures are seen, crowding closely upon the 
mirror, and upon each other, nearly parallel, and of exquisite del- 
icacy. A pleasing and instructive experiment is performed by 
placing a pin perpendicularly upon the paper, and moving it back 
and forth. Its several shadows run along the curves as tangents, 
and reveal at once, for each point, the direction of the rays em- 
ployed in forming it. 
3. Apparatus for experiments on inflection and interference of 
light. 
There is one class of experiments on inflection and interfer- 
ence, which, if the instructor in optics attempts to show them to 
his pupils, must necessarily consume much time. I refer to those 
in which are employed a metallic screen with minute apertures, 
and a magnifier, through which the light, having suffered inflec- 
tion by passing the apertures, falls into the eye of the observer 
beyond. If any considerable variety of combinations is inter- 
posed for acting on the light, much trouble is experienced and 
much time wasted in exchanging one pattern for another. ‘The 
article [ am about to describe is merely a simple contrivance for 
reducing the time and labor of exhibiting this beautiful class of 
optical phenomena. 
In fig. 7, (a, a) represents a wooden ring one foot in diameter, 
one inch anda half wide and one fourth of an inch thick, strength- 
ened by two pieces of the same width and thickness, crossing 
each other perpendicularly in the centre. (6) is an upright flat 
pillar fastened to the heavy base (c), and having a height a little 
greater than the diameter of the ring. A short axis (d) is firmly 
attached to the middle of this pillar, on which the ring is con- 
