428 
removed and perchloric acid remaining. The best 
proportions are: 1 kg. ammonium perchlorate, 
800 ¢.c. water, 600 ¢.c. nitric acid (sp. gr. 1.42), 
200 ¢.e. hydrochlorie acid (sp. gr. 1.2). The hy- 
drochloric acid is added slowly to the boiling solu- 
tion of the other substances, using a reflux con- 
denser. After about one hour the solution is 
evaporated until all nitric and hydrochloric acid 
is removed. The yield is the theoretical. 
CHARLES H. WHITE: A Colorimeter for Rapid 
Work with Widely Varying Standards. 
The standard and the unknown solution are com- 
pared in hollow wedge-shaped prisms of glass, 
placed vertically in a camera behind screens pro- 
vided with a narrow horizontal opening opposite 
the eye for the simultaneous observation of the 
two solutions. The wedges are held in graduated 
carriers, by the vertical movement of which the 
thickness of the section of solution observed is 
varied. The wedge containing the unknown is set 
at the graduation representing the percentage of 
color-producing agent in the standard. The stand- 
ard is then adjusted to match the color of the 
unknown and the percentage in the unknown is 
read directly. 
Irvine Lanemuir: The Dissociation of Hydrogen 
into Atoms at High Temperatures. 
In a previous paper (to be published in the 
Phys. Rev.) it was shown that the laws of heat 
convection in gases are much simpler than has 
been generally supposed. In fact, the problem 
reduces to one of simple conduction through a 
‘film of relatively stationary gas held in place by 
viscosity. The thickness of the film in any gas 
under any given conditions may be calculated 
accurately by simple formulas. If the energy loss 
for a tungsten filament in hydrogen be calculated 
in this way the results agree with observed values 
within the experimental error up to a temperature 
of about 2300° K., but above this show rapidly 
increasing differences. At 3300° K. the observed 
energy loss is over four times the calculated. It 
is assumed that this difference is due to the disso- 
ciation of hydrogen molecules into atoms. The 
variation of the energy loss with the temperature 
and with the pressure agree quantitatively with 
this theory. Simple unforced theoretical considera- 
tions form the basis for a calculation of the heat 
of formation of hydrogen molecules and the per- 
centage dissociation at various temperatures. The 
results are: Heat of formation of H., 125,000 cal. 
Percentage dissociation (1 atmo.): at 2500° K. 
3.9; 2700°, 10.0; 2900°, 22.1; 3100°, 40.4; 3300°, 
SCIENCE 
[N.S. Vou. XXXV. No. 898 
61.4. In the case of the monatomic gas, Hg vapor, 
the calculated energy loss does not show differ- 
ences with the observed values similar to the above. 
Irvine Lanemuir: A Chemically Active Modifica- 
tion of Hydrogen. 
When a tungsten wire is heated to a tempera- 
ture between 1300° and 2500° K. in hydrogen at 
very low pressure (0.001—0.020 mm.) the hydrogen 
slowly disappears. With nitrogen or carbon mon- 
oxide such disappearance never occurs below about 
2200° K. and seems to be an electrical effect, 
while with hydrogen it is purely thermal. There is 
a distinct fatigue effect, but the substitution of 
a new section of wire does not restore the action. 
The hydrogen is not absorbed by the wire, but is 
deposited on the glass, especially where the latter 
is cooled by liquid air. If the wire is allowed to 
cool and the liquid air is then removed, ordinary 
hydrogen is set free which will not recondense 
when the liquid air is replaced. If this gas is 
pumped out and oxygen is admitted the oxygen 
will disappear and in its place a small quantity of 
hydrogen will appear. With platinum and espe- 
cially with palladium wires these effects are much 
more marked. ‘These effects are not due to the 
presence of finely divided metal on the surface of 
the glass. When phosphorus is present on the 
inner surface of the bulb and hydrogen is intro- 
duced and the wire heated to a temperature of 
about 1900° K., the hydrogen disappears and phos- 
phine is formed (apparently the first direct syn- 
thesis of phosphine). These effects are accounted 
for by the theory that the hydrogen dissolves in 
the material of the wire in the atomic condition 
and that some of these atoms leaving the wire, do 
not meet other atoms (because of the low pres- 
sure) but diffuse into the tube cooled by liquid air, 
or become adsorbed by the glass, and thus remain 
in the atomic condition, retaining all of the chem- 
ical activity of the atoms. 
Irvine LANGMUIR: Melting Point of Tungsten. 
The melting point was determined by heating 
pure ductile tungsten wires in pure hydrogen to 
gradually increasing temperatures until the wire 
melted through. The temperature was found from 
the candle power emitted per square mm. of sur- 
face, using a modified form of Rasch’s equation. 
The advantages of an atmosphere of hydrogen are: 
(1) Vaporized metal is carried away by the hy- 
drogen and does not blacken the tube and inter- 
fere with candle-power measurements. (2) The 
energy loss from small wires in hydrogen is nearly 
independent of the diameter of the wire, therefore 
