588 Philippine Journal of Science 1920 
and consequent changes of temperature are prevented by par- 
tially protecting the apparatus with a cylindrical hood of asbestos 
board. This arrangement of the apparatus is of great advan- 
tage because little or no fluctuation of the mercury thread oc- 
curs, and air is excluded by the steam arising from the liquid. 
It is important that the apparatus and its accessories be at room 
temperature (29° C.) before starting the test. 
Normality of alkali. 
0.04 a2 
| abt 
. RAE — wee I [ee een 
| | | t ENS 
4 | 
04 
ron = 
sasee soo SemenQene a: 
ee er 
Lag BS 3 | 
aA es 
6 Ea 
= 
S zo} | 3¢ a i 
oes Sais | Packt 
8 bo | 
s Lon —_{|_}-_+ +— 
S { | i 
° | | Be So 
of | | | 
2 15 i es 
= | | | 
= | | | 
~ i | cee ise Bie S eee 
5 = | | | | 
cs) 2¢ | x 3 moss Coweace Seurens noe 3 
2 -— 
3 DS Se eee Oo ee Soe 
vv | . Pia id ee 
Fak 
- a Lt 
la P+. = a Be meee 
: “teu 
Ee. 
Fic. 3. Effect of normality of alkali on amount of potassium permanganate in onbie eet 
meters of a 0.05 N solution reduced by glucose. Curves 1, 2, 3, 4, 5, and 6 are for eT 
10, 15, 20, and 25 milligrams, respectively. 
PROCESS OF STANDARDIZATION 
In the oxidation process the time of boiling is reckoned from 
the point when the temperature within the heated liquid is 95 
C., rather than from the time when the liquid just begins to 
boil. Numerous trials show that it is practically impossible 
to determine exactly by visual observation when the mixture be- 
gins to boil. There is an easily measurable interval during 
which an observer cannot decide from what point he shall reckon 
the time of boiling. The appearance of bubbles does not usually 
indicate the beginning of boiling because bubbles of gas esc@Pt 
at irregular intervals before real boiling takes place. Slight 
