GASEOUS SPECTRA UNDER HIGH DISPERSION. 421 



most beautifully engraved, map, or picture of the same — yet I have invariably 

 found a mere list of printed numerical spectrum places, by whomsoever issued, 

 to be but very little instructive, — without spending a lamentable amount of 

 time over its interpretation, application and meaning on every occasion of 

 using it ; while it is also not a little expensive to print. Hence I have tried 

 on this occasion to save the cost of figure printing, by throwing the whole 

 burden of what is needed for final results, on the Plates alone. 



Spectroscopic plates of some kind, on account of peculiar virtues of their 

 own for such subjects, must be introduced in some shape, or to some degree. 

 The following characteristic sentence occurs in a British Association Report 

 for 1883, p. 123: — "Three such spectra have been photographed, but without 

 the aid of maps their peculiarities are not capable of description." Wherefore 

 now, by keeping our plates sufficiently large, and with very clear and distinct 

 scales, we may hope that they will not only preserve their own peculiar attri- 

 bute of showing at a single glance the groupings and general bearings of 

 multitudes of lines far better than any other known method, — but they will 

 allow, on close examination and longer inquiry, the numerical places of any 

 particular lines to be read off to nearly as minute a degree of exactitude, as 

 the original observations were capable of giving. 



This good quality would have been absolutely so with the 120 foot records ; 

 and if not quite so with the 40 foot size, that is why I specially request that 

 Mr Heath's drawings shall not be reduced any further, as no smaller scale can 

 pretend either to do justice to the originals, see especially Plates LXXII. 

 and LXXIIL, or impart confidence to those who may use them. To which 

 apology I have only now to add, that the following principles of representation 

 have been strictly followed throughout all the plates of this series. 



Principles of Spectrum Representation. 



Rule 1. The representation is negative, in so far as it shows light by black ; 

 and darkness by white. 



Rule 2. A vertical black line, whether thick or thin, tall or short, on the 

 plate, represents, and is devoted to representing, nothing else than a true 

 spectral line of light, seen and measured as such in the spectroscope, and is 

 considered as precious a result of observation or discovery to the spectroscopist, 

 as a real star is to the Astronomer. 



Rule 3. Different degrees of brightness in the real lines of light in these 

 gaseous spectra, where all lines are necessarily of an equal height, — are 

 represented in the drawings, approximately by different thicknesses of the 

 black lines employed there ; and more exactly, by adding thereto the further 

 method of different heights or depths of the lines so drawn. 



