SECTIONAL TRANSACTIONS.—A. 415 
for various factors; but they added only a few per cent. to the total, and did 
not multiply it by 10 or more. 
The validity of the salinity argument may be tested by two checks—the 
supply of chlorine, and the denudation required to account for the amount of 
sodium; and as shown by Dr. A. Holmes, each of these indicates a much 
longer period than the sodium. 
The supply of chlorine in igneous rocks is quite inadequate to convert their 
sodium into chloride. Most of the sodium chloride in river water is probably 
marine in origin, and only the sodium in the bicarbonate and sulphate is a fresh 
addition to the sea. On this ground the salinity estimate should be approxi- 
mately doubled. Again, to obtain all the sodium in the sea from igneous 
rocks would involve the denudation of improbable volumes of them, and, at 
-the rate usually accepted, the age of the earth should be multiplied three- or 
four-fold. ; 
The fundamental objections to the salinity argument are against (1) its 
assumption that the sea was originally fresh, which paleontological evidence 
renders improbable. The oldest fauna, the Cambrian, has th2 characteristics 
of a marine fauna, and the contrast between the freshwater and marine faunas 
was as sharp in Paleozoic times as it is to-day. (2) Its omission to allow for 
the large supplies of sodium chloride raised from beneath the earth’s surface 
by magmatic waters. (3) Its assumption of uniform denudation. The earth 
has probably undergone deformations that led to alternate periods of quick 
and slow crustal movement ; during the times of repose the surface would have 
been planed down, rivers would have become sluggish and denudation slow. 
As the earth is now under the influence of a time of quick movement, denuda- 
tion is faster than the average. A multiplication of the earth’s age five-fold 
for this difference would not be excessive. 
During quick crustal movement volcanic action would be more powerful, 
the discharge of hydrochloric acid and sodium in hot springs would be in- 
creased ; and as denudation is now acting on land in which sodium chloride has 
been produced in unusual quantities by volcanic action the estimated age of 
the earth must be again extended. The rhythmic acceleration of geological 
processes lengthens the estimates based on sedimentation, but would affect the 
biological argument inversely, since at periods of rapid physical change bio- 
logical change would have been quickened, and thus the occasional abrupt 
introduction of a new fauna does not necessitate so long an interval as has 
been thought. 
The best-known geological estimates of the age of the earth require to be 
multiplied ten- or twenty-fold in order to agree with the physical estimates, 
but this increase is consistent with the geological evidence. 
Prof. A. S. Eppineton, F.R.S.—A study of the Cepheid variable stars affords 
strong evidence that the stars have other sources of energy besides that furnished 
by gravitational contraction. The rate of radiation by 6 Cephei is such that it 
would be necessary for the density to increase 1 per cent. in 40 years in order to 
provide the required energy. The light-change of 5 Cephei is believed to be due to 
a periodicity intrinsic in the star (e.g. pulsation) ; it is clear that such an intrinsic 
period cannot remain unaltered whilst the density changes so rapidly. But 
the observed change of period of 6 Cephei is only 08.08 per annum, or 1 per 
cent. in 58,000 years. The condition of 5 Cephei is thus changing at a rate 
very much slower than that required by the contraction theory. The figures 
suggest that Lord Kelvin’s time-scale should be lengthened 500-fold—at least 
during this stage of the evolution. 
Prof. Sotzas, F.R.S., also took part in the discussion. 
28. Prof. R. A. Sampson, F.R.S.—The Microchronograph. 
The instrument described is in actual use at the Royal Observatory, Edin- 
burgh, for registering clock times of any signal accurately to 0.001 sec., for 
the purpose of examining short-period changes in the relative rates of two 
clocks, or the lag of a controlled clock, or of one part of a piece of apparatus 
with respect to another. It has many other applications. The means employed 
are an adaptation of the oscillograph; this instrument can be so constructed 
as to be completely free from noxious or variable lag of its own. A detector 
