LE SUVOL. Li rMNEO BOE ARYL VPO TEE STS: 67 
It does not appear therefore that there are good grounds for 
assuming a failure of the recognized law of density in such a 
direction as to relieve the great discrepancy shown by the com- 
putations. In any case there is the gravest reason to doubt 
whether it could reach a value represented by a multiplier 
ranging from 140 to 1200, not to say 1800. 
But even if an arbitrary attempt were made to reduce the 
computed moments of momenta to consistency with those of the 
existing system, it is not apparent how it could be attended with 
success and preserve self consistency. The discrepancies are : 
For the Neptunian nebula - - - = BYU WO I 
For the Jovian se - - - - I41 to 1 
Hovthemlerrestialy <6 — - - - = 1208 to I 
(or on the more probable basis) - - 1801 to I 
For the Mercurial nebula - - - - 754 to! 
(or on the more probable basis) — - - 1127 to I 
Now any deviation from the recognized law must be supposed 
to be consistent for analogous conditions. Iftherefore we assume 
such a modification as to bring the moment of momentum of the 
Neptunian nebula into equality with the present moment of 
momentum, we must assume that a similar modification held 
good for all the subsequent stages, either in the same proportion 
or in some systematically increasing or decreasing proportion. 
But the ratios succeed each other in a very arbitrary way, and 
the Neptunian divisor will not bring the others into accord, nor 
will any obvious series of divisors built systematically upon it. 
Were the computation extended to the other nebule, additional 
irreducible irregularities would doubtless appear. 
3. The ratio of masses to momenta.—In the symmetrical 
evolution of a spheroidal nebula by secular cooling, as postu- 
lated in the Laplacian hypothesis, it is reasonable, if not neces- 
sary, to suppose there would be some systematic and rational 
relationship between the masses separated from time to time and 
the moments of momenta of these masses, for the separation was 
due to a common progressive cause, the acceleration of rotation. 
The hypothesis may therefore be tested along these lines. In 
