HVPOTHESES BEARING ON CLIMATIC CHANGES 657 
scant ground for any dissent from the inference that the present 
hydrosphere was then mainly a part of the atmosphere. It does 
not rigorously follow, however, that this was so. Hypothesis may 
go so far as to attribute much of the subsequent ocean and 
atmosphere to vapors thrown out of the molten magma as it 
cooled and to vapors gathered from space since. But I venture 
to question the supposed original molten state. While making 
no claim to disprove it, 1] doubt whether it rests on sufficiently 
solid theoretical grounds to justify the assumptions so unhesita- 
tingly built upon it. There is still some ground to doubt the 
nebular hypothesis and to entertain some of the various phases 
of the meteoroidal hypotheses. The nebular hypothesis corre- 
lates a wonderful array of remarkable facts and has gained a 
profound hold upon the convictions of the scientific world, yet 
some of its great pillars of support have recently weakened or 
have fallen away entirely. Of the 5000 known nebule to which 
we naturally look for analogy very few, if indeed any, strictly 
interpreted, exemplify in a clear and decisive manner the sys- 
tematic annular evolution postulated by Laplace. The pho- 
tographs of the nebula of Andromeda, that were hailed with such 
delight on their first appearance as exemplifying the Laplacean 
hypothesis, appear upon more critical study to support it only in 
vague and general terms, if indeed they lend it support at all. 
The Saturnian rings, the trite source of illustration and analogy, 
prove under the test of the spectroscope to be formed of discrete 
solid particles, and not of gas, and the investigations of Roche 
have put a new phase on their theory. While in their form they 
tally with the annular hypothesis they do not support its gas- 
eous phase, if indeed they lend it any support at all. But our 
chief interest is not in the general theory, but in the special 
inferences drawn from it respecting the early stages of the earth. 
Let us assume the possible or, if you please, probable truthful- 
ness of the nebular hypothesis so far as the separation of an earth- 
moon ring from the shrinking sun is concerned. Do the subse- 
quent steps commonly postulated logically follow ? 
The vast radiating surface of such a ring, its attenuated 
