440 • Colours of Stars. 



The atmosphere also introduces a certain amount of occa- 

 sional deception. It is obvious that a degree of haze which 

 gives a red or yellow tinge to the sun by day must produce 

 the same effect on white stars by night ; and on this account 

 the colours estimated on different nights might be found to 

 vary, and even on the same night at different altitudes above 

 the horizon. Hence the tints of low- culminating stars can 

 seldom be satisfactorily determined, even if we could eliminate 

 the effect of refraction, which interferes again in its own way, 

 converting circular discs into lengthened and parti-coloured 

 spectra, and sometimes, as Smyth observes, making " a large 

 star of a white colour really appear like a blue and red hand- 

 kerchief fluttering in the wind; the blue and red about as 

 intense and decided as they could well be." The lowest 10° or 

 15° of the visible heavens are on this account commonly con- 

 demned by astronomers as useless ; but Herschel I. found 

 traces of this prismatic effect even as high as Regidiis, and 

 observed that from this cause a star was not always best seen 

 in the centre of the field, there being a position where the 

 prismatic error of rays passing obliquely through the eye-lens 

 may, in some measure, correct that arising from atmospheric 

 refraction. Struve, in later days, traced prismatic effects from 

 this cause to 30° and even 45° froni the horizon. 



Some care should be taken as to the standard and nomen- 

 clature of colour, as discrepancies may arise from carelessness 

 or inattention on this head. There is unquestionably a natural 

 or intrinsic standard of colour in the primary tints of the spec- 

 trum, but they do not come before us in an unmingled form 

 decidedly or frequently enough to be impressive on the 

 memory ; practically speaking, each blue that we see may be 

 thought greener or more purple, if compared with other shades 

 verging more to purple or green than itself; and so our ideas of 

 yellow oscillate through a considerable interval between green 

 and orange; and red has many variations between orange and 

 purple. Besides this, the language of many persons is habitu- 

 ally vague ; and in the cast" especially of mingled tints, different 

 names might bo given by different persons to the same colour. 

 To obviate these causes of uncertainty Smyth's excellent sug- 

 gestion should be adopted, of referring all hues to correspond- 

 ing water-colour pigments, whorb the definite name admits of 

 no question, provided only tho memory of the eye may bo 

 depended upon. 



But beeidei these comparatively external sources of error, 

 \ve have to observe that tho judgment of tho eye itsolf may be 

 easily led ;istr;iy. To the difference which, as has been already 

 intimated, may exist between the perceptions of different indi- 

 viduals, wo havo to add those Avhich may casually arise in the 



