Apeil 1, 1910] 



SCIENCE 



511 



formity throughout the world, all the figures 

 lying between a fourth and a fifth of the total 

 area. He too was impressed by the smallness 

 of the variation in the relative areas of 

 Archean exposures. 



It seems well established that at the present 

 day the relative area of y/A lies between two 

 tenths and three tenths. The values for the 

 constant c, the ultimate sodium accumulation, 

 Amc, and the present age, t, as computed from 

 the formula are given in the following table 

 for these extremes and also for y/A = 1/4. 



ylA 0.20 0.25 030 



23.636 X 10' 31.515 X 10« 40.519 X 10' 

 Amc N 5/4 N 4/3 N 10/7 



t 38.0 X 10" 43.7 X 10" 48.8 X 10= 



It can also be computed from the formulas 

 at what rate the area of massive rocks is di- 

 minishing. This is expressed by dy/dt = 

 — y/c. Substituting for y its value in terms 

 of A and taking A at 134.38 X 10° square kilo- 

 meters gives the mean annual net decrement 

 for each of the three cases at almost exactly 

 one square kilometer in spite of additions due 

 to vulcanism, a result which is certainly not 

 startling. 



Mr. F. W. Clarke" has shown that a shell of 

 average igneous rock enveloping the globe and 

 2,225 feet (678.2 meters) thick would furnish 

 all the salt of the ocean. The ultimate thick- 

 ness of the decomposed shells corresponding 

 to the three values of y/A would be propor- 

 tional to Amc, that is, 848, 904 and 969 meters, 

 respectively, so that Clarke's shell would never 

 assume improbable dimensions, or exceed three 

 fifths of a mile. 



Much the weakest point in this speculation 

 seems to me to be the assumed constancy of 

 the land area. This has assuredly fluctuated, 

 yet when shallow seas flooded portions of the 

 continents, marine denudation took the place 

 of erosion in part, at least, and was possibly an 

 equivalent. A is taken as constant only be- 

 cause there seems no way at present in which 

 its variations can be rationally represented. 

 The present marine denudation is offset to 

 some extent by wind-borne or cyclic salt. It 



° " Data of Geochemistry," 1908, p. 28. 



seems to me needless to consider the sodium 

 content of sedimentaries as a source of supply 

 for the ocean. Limestones contain a mere 

 trace of sodium. Shales — ^which are the prev- 

 alent detrital rocks — (including clay, clay- 

 slate and phyllite) contain sodium, but seem- 

 ingly in a stable form, since ancient phyllites 

 and modern clays are indistinguishable by 

 their sodium content. Sandstones do not con- 

 tain enough sodium to affect the problem of 

 the earth's age, considering the greater uncer- 

 tainties. Possibly some massives underlying 

 thin layers of sedimentaries yield a little 

 sodium, but, per contra, areas properly mapped 

 as Archean or massive are in many localities 

 protected by saprolite. This is true locally 

 even in glaciated Alaska. Considering such 

 protection, I believe the effective value of y/A 

 to be nearer a fifth than a quarter. 



Possibly it may be worth while to refer to 

 the evident fact that if the descending ex- 

 ponential properly represents the history of 

 the accumulation of sodium in the ocean, this 

 became highly saline much earlier than on Mr. 

 Joly's theory. Thus when the earth was half 

 its present age the law of linear increment 

 would of course imply that there was half as 

 much salt in the ocean as there is now, while 

 if the exponential relation holds good and 

 y^A/5, the ocean at that epoch contained 

 seven tenths of the present amount. The 

 fauna of the Paleozoic indicates a salt-water 

 habitat. If the deep was then as briny as it 

 now is, it must have taken in a vast amount 

 of juvenile water in the mean time on either 

 hypothesis. 



The annual increment of N or the quantity 

 here called my is susceptible of improved de- 

 termination, as every one has recognized. Mr. 

 Clarke is now engaged on a discussion of this 

 subject based on far more extensive material 

 than was at Mr. Joly's disposal, and his re- 

 sults wiU be available in a few months. When 

 they are known, it will take only a few minutes 

 to recompute the age of the earth on the hy- 

 pothesis here discussed. 



The foregoing speculation is based on the 

 assumption that the area of sodiferous rocks 

 has diminished by a constant proportion (1/c) 



