173 



therefore, as it is known that prismatic colours exist even in 

 these minute refractions, a corrected glass should reveal their 

 existence; different kinds of glass form an interesting 

 study. 



The beading of a great many microscopic objects, as 

 diatoms and scales, is prismatic and more or less possesses 

 dispersive powers ; and they are the most beautifully and 

 distinctly defined only when they begin to flash with the 

 rays of the emerald and ruby. When these colours cannot 

 be seen, in general their definition is defective. 



But a still higher order of test is to be found in the 

 observation of artificial double stars or triplets. Quicksilver 

 carefully cleaned by pressure through several folds of soft 

 leather, scattered on a black surface, may readily be divided 

 into minute particles, and those about the 1-lOOOth of 

 an inch form beautiful images. To produce double micro- 

 scopic stars, two round apertures should shed brilliant 

 pencils upon the globule ; then, according to their distance 

 apart, and from the globule, their images will appear to 

 approach or recede from each other. At a certain point 

 and angular distance from the minute, mirror these double 

 stars coalesce. This effect measures the aberration. These 

 apertures, arranged at equilateral distance, form beautiful 

 triplets. 



According to the investigation of my friend Professor 

 Bashforth, who has kindly furnished me with a tracing of 

 the curve taken by minute globules of mercury, the reflecting 

 surface becomes more nearly spherical as the particle dimin- 

 ishes, and the upper surface forms the most accurate images. 

 The aberration is better determined, therefore, on the upper 

 surface of the mercury. As a rough approximation, two 

 small lamps placed one foot distance from the globule, and 

 two inches apart, may be first experimented upon with re- 

 flecting particles arranged from the 1-lOOth to the 1-lOOOth 

 of an inch. The smallest globules form the more severe test, 

 and the double stars will be easily divided as larger particles 

 are selected. 



Phenomena observed. 



Generally diffraction rings coalesce and become oval as the 

 drop is smaller. The shape of the flame is obscured; radiat- 

 ing rays ensparkle the images and conceal the definition. 



Reduce the aperture of the objective. The images become 

 fainter, but more distinct. Increase the distance between the 

 front glasses of the objective, and so acljust for an uncovered 



