on the Optical Phenomena of the Atmosphere. 11 



however, such investigations are made more with a view to de- 

 termine the relative brightness of the stars, compared with each 

 other, than to ascertain the alterations in the intensity of their 

 light at various zenith distances. For the more direct investi- 

 gation of the transparency of the atmosphere, Saussure's diapha- 

 nometer seemed to us to be peculiarly suitable*. This con- 

 sists of two white circles, in the middle of which are placed black 

 discs of various diameters (fig. 2) . The greater circle (aa) has 

 a diameter of 2 P. F. ; the black disc within it (bb), a diameter 

 of 1 foot ; the entire circle is surrounded by a green rim (c) . In 

 one corner of the latter a smaller circle is similarly placed ; its 

 diameter is 2 inches, and the diameter of the black disc within 

 it 1 inch. Saussure's instrument was double the size of ours. 

 We chose these smaller dimensions, because we feared that on 

 the higher summits it would be difficult to obtain the distances 

 necessary to cause the disappearance of the larger discs ; more 

 difficult still would have been the accurate measurement of such 

 distances at these elevations. We sometimes encountered, even 

 with our smaller instruments, difficulties of this nature. 



In the application, the experimenter recedes from the discs 

 thus placed together, until the black centre of the smaller one 

 can no longer be distinguished from the surrounding surface. 

 The same process is repeated with the large disc. If the air were 

 perfectly transparent, the angles under which both discs disap- 

 pear would be equal ; and as the angles in this case are very 

 small, it may be assumed that the distances from the centre and 

 from the borders of the discs are the same. The respective di- 

 stances are then in the ratio of the diameters of the two discs. 

 The diaphanometer of Saussure is indeed no absolute standard 

 for the transparency of the atmosphere ; the eye of the observer, 

 the intensity of the colour of the instrument, and the manner 

 in which it is set up, all have an influence upon the results. By 

 using proper caution, however, the error may be greatly lessened ; 

 and the differences at various elevations are so considerable, that 

 its influence on this account is still further diminished. As the 

 observer's eye seems to be the most variable among these sub- 

 jective sources of error, a few physiological remarks on this sub- 

 ject may assist in the forming of a judgement as to the following 

 experiments f- When we recede from the discs until the black 

 circle vanishes, the angle under which the instrument is then 

 seen is so small, that the contiguous portions of the retina are 

 impressed at the same time by the black and white portions of 

 the surface; the impressions thus unite to form a dull gray 



* Mem. de Turin, vol. iv. 1/88 and 1789, p. 425-440. 

 t Compare Volkmann's beautiful memoir Ueber das Sehen in Rudolph 

 Wagner's Physiological Dictionary, Lit. S. p. 263-351, 



