664 T. C. CHAMBERLIN 
truth. The number of collisions also increases with the density 
of the gas. In the supposed case of an atmosphere containing 
all the water of the globe, the density would perhaps be 300 
times the standard density. In the upper regions the density 
would be low and ;4, of the standard density may be taken as 
a representative of the conditions there. Making the assump- 
tion ‘that ithe collisions (of waterivapon are VOL!) perm secs mune 
periods required for all the molecules, on an average, to acquire 
the parabolic velocity of the earth would be as follows: 
At At qty At 300 times 
Temperature Standard Pressure Standard Pressure Standard Pressure 
1000° 1.7 X 107° years io XX WO? Ears 5.7 X 1073 years 
I500° 8) < UO! TODS TOM AS Beis SM noe 
2000° AKI «& AK Ie iol KTE® & 
3000° 33 years 3300 years 4o days 
4000° 1030 seconds 28.5 hours 3.4 seconds 
Under the current hypothesis of a molten earth derived from 
a gaseous one the temperature of the atmosphere would probably 
exceed 4ooo C. during the stages of condensation of the 
refractory material of the earth from the form of a gas to the 
form of a liquid. From this fervid stage the temperature would 
fall to 2000° C., or below before the crust would begin to form 
and the external louid condition cease. The temperatures of a 
liquid earth may therefore be assumed to range from 4000° C. 
to 2000° C. or below, and the figures of the preceding table may 
be interpreted on this basis. 
If the question were simply the acquisition of molecular 
velocities at the surface of the liquid mass greater than the 
parabolic velocity of the earth at that point within an available 
length of time, it would appear that the retention of the water 
vapor would be put in serious jeopardy during the hotter stages, 
but that it might survive the cooler ones in large part if it 
reached them, unless they were very prolonged. But there are 
other considerations to be taken into account. Under the most 
favorable conditions only a part of the molecules which attain a 
speed beyond the limital velocity of the earth’s control would 
