350 Molecules, Ultimates, Atoms, and Waves. [July, 



neous impulse, so that the medium is perfectly elastic. 

 Minute waves of various degrees of length are supposed to be 

 propagated through this elastic ether, when any of its par- 

 ticles are disturbed. These waves manifest their existence 

 in the phenomena of heat, light, actinism, and fluorescence 

 — the longer waves in heat, the shorter in actinism and 

 fluorescence, the intermediate in light. 



According to the most recent determinations, the speed 

 with which these waves traverse the ether is 185,000 miles 

 in a second, or nearly 11,721-millionths of an 'inch in the 

 billionth of a second — this last being the most convenient 

 form in which to state the speed. 



Great labour and skill have been bestowed on the measure- 

 ment of the lengths of those minute waves. Fraiinhofer, 

 who first discovered the fixed lines which bear his name, 

 measured the wave-lengths corresponding to the lines which 

 he named b, c, d, e, f, and h, leaving behind him two 

 sets of measurements, differing slightly from each other. 

 More recent observations with more delicate appliances than 

 were at his command, have shown the lines D, e, and H to 

 be double. Professor Angstrom of Upsal, availing himself 

 of the best instrumental means that could be obtained, has 

 re-measured the wave-lengths of all the principal lines from 

 A to the more refracted h, inclusive. The results he has 

 published along with an elaborate atlas of the spectrum, in 

 which the lines are laid down according to their wave- 

 lengths. The spectrum represented in this atlas is the dif- 

 fracted or normal solar spectrum, obtained from a system 

 of very fine equidistant lines. In all spectra produced by 

 prisms the fixed lines are displaced relatively to each other, 

 so that their true wave-lengths cannot be determined from 

 any such spectra. This displacement is termed the irra- 

 tionality of the spectrum — a phenomenon discussed at large 

 in a paper inserted in the " Philosophical Magazine" for 

 i860, pp. 165, 263, and 364. This displacement is different 

 in every different medium, and even in different sorts of 

 glass ; while it is more or less affected also by the tempera- 

 ture of the medium. Thus the fixed lines will have different 

 positions relatively to each other in different spectroscopes 

 and at different temperatures, while the greater the number 

 of prisms the greater will be this source of error — a circum- 

 stance which has been rather overlooked in recent observa- 

 tions on spectrum analysis. 



A plan of a portion of the solar spectrum has also been 

 published by M. Kirchhoff, and of the remainder by Messrs. 

 Hofmann, Angstrom, and Thalen— all drawn to the same 



