238 Intelligence and Miscellaneous Articles. 



trodes for the magnesium ones, the lines of the iron-spectrum (espe- 

 cially the one situated between the two doubled magnesium-lines) 

 photographed themselves with perfect distinctness , there was there- 

 fore no error on this side. 



I then concluded that the doubling of the lines was a true reversal — 

 the analogue, in the invisible portion of the spectrum, of the reversal 

 of the line D, obtained by M. Fizeau on placing between the two 

 carbon points of the electric lamp a fragment of sodium. 



No hesitation would have been possible if the three lines had been 

 doubled simultaneously; but it seemed strange that one of them 

 should escape this modification. I successively took twenty-two 

 plates ; and in none did the most refrangible line appear doubled. 



I then recollected an observation of Father Secchi on the spec- 

 trum-analysis of a solar spot (May 1869)*, in which the able astro- 

 nomer relates his having witnessed the reversal of one only (the 

 least refrangible) of the three magnesium-lines which constitute b. 

 This near resemblance induced me to persevere ; and I was fortu- 

 nate enough to produce at will, on the same plate, the reversed or 

 the normal lines. One can even compare the positions of the normal 

 and the reversed lines, just as the spectrum of an artificial light is 

 compared with that of the sun. It is only necessary to cause the 

 spark of a powerful induction-apparatus to issue between the mag- 

 nesium electrodes, covering one half of the slit of the spectroscope, 

 then the spark from a very feeble apparatus f, prolonged a sufficient 

 time, and to cover the other half of the slit. The exact coincidence of 

 the normal and reversed lines is easily ascertained. 



The reversal in the ultra-violet portion being beyond doubt, it was 

 necessary to obtain it also in the visible region of the spectrum. 

 The spark of the large induction-coil, even with 24 couples arranged 

 as 12 with a double surface, produced no effect on the magnesium- 

 line b ; but the experiment succeeded easily with the voltaic arc of a 

 pile of 50 couples. It can be effected thus : — For the positive pole a 

 disk of carbon 6-8 centims. in diameter is taken, in which are sunk 

 little capsules ; in each of these is placed a fragment of metal ; then 

 the negative carbon is lowered, and then raised as soon as the spark 

 has issued, so as to obtain an arc of 4 or 5 rnillims. With the aid of 

 a lens the image of the arc is projected on the slit of the spectroscope. 

 In the case of magnesium the triple line b appears bright and clean : 

 it is to be placed exactly in position. Then the upper carbon is 

 lowered by degrees ; the lines widen, become blurred, and very soon 

 a very fine black line appears on the least refrangible ; if the ap- 

 proximation of the carbons be continued, the second and at last the 

 third line will be reversed. Here, then, is the verification of the 

 phenomenon observed photographically, and the artificial reproduc- 

 tion, although in the opposite direction, of that described by Father 

 Secchi. 



* Coynptes Rendus, vol. lxviii. p. 1243. The author adds that the line 

 C " appeared sometimes doubled ; but," he says, " I attribute this pecu- 

 liarity to the movemeDt of the terrestrial atmosphere." May not this also 

 have been a true reversal ? 



f A medium induction-coil, 3 or 4 couples, and a single Leyden jar of 

 about a litre capacity. 



