Intelligence and Miscellaneous xir tides. 157 



or in a current of illuminating-gas. The oxidizing layer which en- 

 veloped the flame is thus rendered reducing ; and we observe that 

 a ilame a little reducing augments the light of lithium and dimi- 

 nishes that of sodium. 



3. It was first ascertained that solutions diluted to jJ-q- are con- 

 verted to powder and carried along by the gaseous mixture in the 

 same quantity, whatever be the substance dissolved. Tor that 

 purpose a crystal of nitrate of calcium is added to a portion of a 

 solution of nitrate of strontium ; the liquid gives the spectrum of 

 strontium with the same brightness as the original solution. The 

 experiment repeated with other salts gives the same results. It is 

 therefore sufficient, in order to compare salts having the same base 

 and different acids, to prepare dilute solutions of them containing 

 the same quantity of metal per litre and submit them to experi- 

 ment. Each salt is compared with the nitrate by pulverizing each 

 of the two solutions alternately, and thus making four or six cross 

 experiments. Here is a series of measurements made with a 

 slightly reducing flame, at 1 centim. above the orifice : — 



Ni- ^^\^ Chlo- Bro- t ti ^hos- Carbo- ^ , , 

 , , sul- .-, -J Iodide. 1 , ^ Tartrate, 



trate. , , ride. mide. phate. nate. 



Sodium .. 1 1001 1-010 0-991 0'990 1-013 0-998 1-014 

 Lithium.. 1 1-003 0992 1-008 1-009 0-996 ... 0-992 



The differences are of the order of possible errors ; therefore the 

 brightness of the spectrum, of these two metals, does not depend 

 on the element with which the metal was combined. Nor has a 

 slight excess of acid any more influence ; but with a notable 

 quantity of chlorhydric acid, bromine, &c. the brightness is 

 diminished. 



Calcium and strontium present some remarkable peculiarities. 

 Their acid phosphates give but a very feeble spectrum. Nitrate o£ 

 calcium gives a more feeble spectrum than the other salts — chloride, 

 bromide, iodide, acetate*; the ratio varies from 1*47 to 1-93, and 

 is the same for the principal bands. Moreover, for one and the 

 same salt, the ratio varies with the conditions of the experiment. 

 It must be remarked that the lines proper to the chloride and the 

 other non-decomposed salts are not visible ; all the salts give the 

 same spectrum. 



The deportment of the salts of strontium is the same, but with 

 less variations. The brightness of the spectrum being equal to 1 

 for the nitrate, it varies from 1*2 to 1*5 for the other salts. 



4. From what we have just seen, the quantities of salts intro- 

 duced into the flame are in the same ratio as the richness of the 

 solutions employed. Thus, when the quantity of salt introduced 



* It was ascertained that these diflferences do not depend on unequal 

 pulverization, by dissolving in each of the liquids the same quantity of a 

 the lines of these metals had the same bright- 



