Miscellanies. 365 
inch in diameter, divided into two hundred parts of equal capacity, 
and supplied with a lip, so that a liquid can be poured from it with- 
out being spilled. In conducting this part of the process, the opera- 
tor will perceive two odors :—at first, the characteristic odor of chlo- 
rine, accompanied with the peculiar irritation of that gas; and sub- 
sequently, an agreeable, somewhat aromatic odor, unattended with 
ihe slightest irritation. The object is to add exactly so much solution 
of iron as suffices to destroy the former of these odors, without at- 
tempting to remove the latter; a point which, with a little practice, 
may be readily attained. The whole of the iron is thus brought into 
the state of peroxide. 
The first trial is generally accompanied with some loss of chlorine, 
and should only be used as a guide to a second and more precise ex- 
periment. Accordingly, a weighed portion of the same ore is dis- 
solved, and the chlorine collected as before, except that the solution 
of green vitriol, in quantity rather less than sufficient, is at once in- 
troduced into theinverted tube and capsule. A more ready and per- 
fect absorption of the chlorine is thus effected, and the subsequent 
addition of a small quantity of sulphate of iron suffices for comple- 
ting the process. 
The principal sources of error in this method are the two follow- 
ing :—loss of chlorine, by smelling repeatedly, and exposure to the 
air when the gas is absorbed by pure water; and oxidation by the 
air when the absorption is made directly by means of the solution of 
iron. : 
The small flask and inverted tube are apt to retain the odor of 
chlorine, and should, therefore, be rinsed out with the absorbing liquid. 
t should be remembered, also, that a given quantity of chlorine will 
emit a more or less distinct odor, according as it is more or less dilu- 
ted. But by operating always in the same manner, and employing 
such weights of different ores, that equa] quantities of the solution 
may contain nearly equal quantities of chlorine, it is easy to be in- 
dependent of these errors of manipulation, by causing them to affect 
each experiment to the same degree. 
It will accordingly be found, with a little practice, that results of 
Surprising uniformity may be thus obtained ; and even the constitution 
pure oxides of manganese may be ascertained by this method, 
almost with the same accuracy as by directly determining the quan- 
tity of oxygen.—Jour. of Roy. Inst. 1. 293. 
