530 C. PIAZZI SMYTH ON TPIE 



So much then for the strength of the illumination of the back-ground of 

 continuous spectrum light, whereby alone any Fraunhofer lines can be 

 distinguished at all. But amongst Fraunhofer lines themselves, there are very 

 great differences ; and while there is nothing so easily imitated or represented 

 by merely drawing a sharp, simple line with a ruling pen and black ink 

 on white paper, as a strong, well-defined, Fraunhofer line, — there are all sorts 

 of deviations from such an ideal in the course of the Solar Spectrum. For 

 there, every experienced observer knows so abundantly, that besides lines 

 thick and lines thin, there are lines of various degrees of paleness, and of 

 various degrees of sharpness, as well as sometimes of even extravagant haziness ; 

 and all these are physical facts which should be expressed to some extent, — 

 though minute accuracy does not stand with them on the same high level of 

 importance as with accuracy of place in the horizontal scale. 



Hence with those two great "Dioscuri" leaders of modern Spectroscopy, 

 Kirchoff and Angstrom, the former's map is not so popular now as it once was, 

 — and partly because the method he adopted of indicating both paleness and 

 thinness of some lines, by printing them in from pale tint stones amongst black 

 lines previously printed from another stone, both sacrificed accuracy of place, 

 and produced very ultra ideas of colour as well as paleness. While Angstrom's 

 map, which printed every line from one stone, with one inking, and sought to 

 give thinness and paleness by exquisitely fine engraving only, — still holds its 

 own among Spectroscopists with remarkable power and tenacity. 



Yet his is not a perfect method, for it cannot show such an undoubted 

 existence as is occasionally met with, in the shape of a h'oad Fraunhofer line 

 of pale material. While partly to make up for that graphical weakness as well 

 as some others, the plan of pure engraving has been made the parent of a most 

 Avidely followed, yet distinctly vicious, system of representing shade, especially 

 pale shade, in the spectrum by thin, close, vertical and parallel lines. 



Now some shady, nebulous bands to inferior spectroscopes, do undoubtedly 

 resolve themselves into thin lines in a better instrument, but some of them do 

 not, and ought not, in any spectroscope whatever; and even with those that do, 

 the question of importance is to determine, into how many lines, and at what 

 distances apart ? Wherefore for an observer who has seen nothing whatever 

 but a hazy, nebulous shade, to represent that in his map by clean, distinct lines 

 of his own invention, drawn just as they would, or ought to, have been drawn, 

 if he had seen veritable Fraunhofer lines in that place, is a species of wilful, 

 scientific misleading which should be tolerated no longer. 



I have had no scruple therefore in my own maps, and also in my copies of 

 anything really important from other men's maps, in adopting as a governing 

 principle for representing both paleness, and degrees of haziness in the spectrum, 

 a symbolic method, not only easy of execution but perfectly impossible to be 



