98 cosmos. 



The so-called relations of the magnitude cf the fixed star* 

 as given in our catalogues and maps of the stars, sometimes 

 indicate as of simultaneous occurrence that which belongs to 

 very different periods of cosmical' alterations of light. Tha 

 order of the letters which, since the beginning of the seven- 

 teenth century, have been added to the stars in the general- 

 ly consulted TJranometria Bayeri, are not, as was long sup- 

 posed, certain indications of these alterations of light. Arge- 

 lander has ably shown that the relative brightness of the 

 stars can not be inferred from the alphabetical order of the 

 letters, and that Bayer was influenced in his choice of these 

 letters by the form and direction of the constellations.* 



fleeted rings are complementary to those of the transmitted rings ; these 

 two series of rings neutralize one another when the two lights by which 

 they are formed, and which fall simultaneously on the two lenses, are 

 equal. 



" In the contrary case, we meet with traces of reflected or transmit- 

 ted rings, according as the light by which the former are produced ia 

 stronger or fainter than that from which the latter are formed. It ia 

 only in this manner that colored rings can ha said to come into play in 

 those photometric measurements to which I havs directed my atten- 

 tion." 



(b.) Cyanometer. 



" My cyanometer is an extension of my polariscope. This latter in- 

 strument, as you know, consists of a tube closed at one end by a plate 

 of rock crystal, cut perpendicular to its axis, and 5 millimetres in thick- 

 ness ; and of a double refracting prism placed near the part to which 

 the eye is applied. Among the varied colors yielded by this apparatus 

 when it is traversed by polarized light and the prism turns on itself, wo 

 fortunately find a shade of azure. This blue, which is very faint, tha. 

 is to say, mixed with a large quantity of white when the light is almos* 

 neutral, gradually increases in intensity in proportion to the quantity oi 

 polarized rays which enter the instrument. 



14 Let us suppose the polariscope directed toward a sheet of w T hite 

 paper, and that between this paper and the plate of rock crystal there 

 is a pile of glass plates capable of being variously inclined, by which 

 means the illuminating light of the paper would be more or less polar- 

 ized ; the blue color yielded by the instrument will go on increasing 

 with the inclination of the pile ; and the process must be continued un- 

 til the color appears of the same intensity with the region of the atmos- 

 phere whose cyanometrical tingle is to be determined, and which ia 

 seen by the naked eye in the immediate vicinity of the instrument. 

 The amount of this color is given by the inclination of the pile ; and if 

 this portion of the apparatus consist of the same number of plates formed 

 of the same kind of glass, observations made at different places may 

 readily be compared together." 



* Argelander, De fide Uranometrice Bayeri, 1842, p. 14-23. "In ea« 

 dem classe littera prior majorem splendorem nuUo modo indicat" ($ 

 9). Bayer did not, therefore, show that the light of Castor was raore 

 intense ? n 1603 than that of Pollux. 



