48 Dr BREWSTER on the Construction of Poll/zonal Lenses 



tal purposes, where the eye alone is to be the judge, the ope- 

 rator's eye would be sufficiently accurate for directing such a 

 process ; but when we consider, that the object is to reflect di- 

 vergent rays into a beam of light, which is required to preserve 

 its parallelism and its density for 30 or 40 miles, we cannot but 

 wonder that such inadequate means should have been so long 

 employed to produce this effect. 



Even if the light in the focus of the hammered reflector were 

 a mathematical point, the most favourable of all suppositions, it 

 would, after reflection, be thrown into divergent pencils a short 

 way beyond the mouth of the reflector, and the resulting column 

 would soon cease to preserve its density and its parallelism. 



4. On the Disadvantages arising from the size of the Arga/nd- 

 burner. As the argand-burner now in use cannot admit of di- 

 minution, it may seem strange that its magnitude should be 

 ranked among the disadvantages of the present system. If a 

 burner an inch in diameter were placed in the focus of a lens, or 

 even in the focus of a large spherical mirror, it would not produce 

 the same imperfections in the reflected column as it does in the 

 focus of the hammered paraboloid. In a reflector 2 feet in dia- 

 meter, the circumference of the wick is only 3|- inches from the 

 apex C of the curve ; but as the glass-chimney which surrounds 

 the flame is nearly 2 inches in diameter, and as the rays from 

 the wick are refracted by the irregularities of this glass, we may 

 safely assume that the virtual diameter of the mass of light, 

 which is the source of illumination, is nearly 2 inches. Now, as 

 the nearest point of the luminous body is only three inches from 

 the apex C, while the most remote is Jive inches, it is manifest, 

 that no parabolic curve can reflect such pencils into a parallel 

 beam ; nay, it is quite clear, that these two pencils must quit the 

 reflector in a divergent state, and must, at no great distance, be 

 thrown into the sea, or scattered upwards in the atmosphere. 



