44 C. PIAZZI SMYTH ON THE LITTLE b GROUP OF LINES, ETC. 



copper or steel plates ; — or shall we adopt a short, easy, symbolic method by which in the tenth 

 of a second we can make a mark anywhere, on paper as well as copper-plate, signifying merely, 

 or standing for, such and such an artistic, or realistic, effect ? 



A splendid example of the former method is to be seen in M. Coknu's fine engravings of 

 the ultra-violet portion of the Solar Spectrum ; for there, thin lines are successfully repre- 

 sented of every degree of shade from lightest to darkest ; but without the mechanical means by 

 which they are executed appearing to the eye, until reinforced by a powerful magnifying glass ; 

 and they are then seen to be the effect of minute dots more or less closely packed, and without 

 ever recurring to " the vulgar expedient " of ruling an actual line, in one uniform degree of 

 blackness. When any person donates the science of his time with such a Solar spectrum map, 

 whether in whole or part, the public ought to be very grateful to him ; and it is to be hoped 

 they are so, in this case, to M. Cornu, the distinguished Parisian scientist, for his unmatched 

 portraiture of " the region of fluorescence." 



But the daily work of the world requires an interim employment of some easier method, 

 something akin to writing by the letters of the alphabet as signifying sounds, in place of paint- 

 ing pictures of the things intended. One method already extensively in use and much to be 

 commeuded, is to express the different degrees of darkness of a Fraunhofer line, whatever 

 the breadth, by different heights or depths of the line ; and this it will be seen I have availed 

 myself 'of in several instances. 



The same principle may be applied to the shadings by which bands are represented in the 

 Spectrum. Shadings of some kind are necessary there with the very faint bands, to prevent 

 straining the symbol too far ; which would result, if we had only the device of shortening the 

 height of a band in positive black ink, to represent an ultra faintness of shade. Such shade 

 being, in reality, at the telescope, just as high necessarily as any of the blackest lines, 

 because they are all reproductions of one and the same slit in front of the spectroscope. 



Faint bands therefore have long since been generally represented by a shading of thin 

 parallel lines ; and the method is unequivocal when the lines are ruled in any direction except 

 that, in which they might be repetitions of the slit of the Spectroscope, or stand for separate, 

 independent, thin Fraunhofer lines. 



Hence I have represented the hazy borders of the b lines, by either horizontal, or 45° inclined 

 lines, and no spectroscopist will take them, or the knots in them, as anything else than 

 symbols of shade. It is necessary too, to be very particular on this point, because the Royal 

 Society, London, in both its Philosophical Transactions and Proceedings, has most pertinaciously 

 set, and is still setting, the opposite example of representing shade in the spectrum, by thin lines 

 ruled parallel to each other and in the direction of the slit ; so that when many and many a 

 close double, or treble, or quadruple Spectral line appears in a Eoyal Society engraved 

 Spectrum plate, it may mean, either that their observer did see a double, or treble, or quad- 

 ruple line in that place ; or, that he only saw a faint, unimportant haze. 



Finally in our present plate, one mere difficulty has to be compassed ; for it is to be an 

 historic memento, from the earliest spectroscopic times to the latest. In my last year's publica- 

 tion, "Madeira Spectroscopic" I attempted to meet a similar case, by representing every 

 observation, of whatever date, on one and the same scale. But I have been told that some 

 persons do not like the necessarily resulting effect of that plan, in so far as it makes the earlier 

 observations, taken with very small dispersive power, look colossal in coarseness. In the 

 present case therefore, I have decreased that appearance by adopting a smaller and smaller 

 scale for every earlier Spectrum view ; but so arranging them one over the other, each with 

 its own sized numerical graduation above it, that I trust there will be the least difficulty and 

 the most satisfaction practically possible, in comparing details of the one with the other, and 

 fixing the dates when real advances of observation — knowledge were made. 



C. P. S. 



