30 CLARKE— STATISTICAL METHOD 



[April i8 



In another calculation, Reade has estimated that the limestones 

 are equal in bulk to a zone 528 feet thick completely enveloping the 

 globe. Then, from the rate at which lime salts are carried from 

 the land into the sea, he computes the time required to form the 

 limestones as equal to 600,000,000 years. I may note here, in pass- 

 ing, that if my own estimate of the mass of the limestones is cor- 

 rect, this quantity should be divided by four, giving 150,000,000 

 years as the time needed for their formation. I do not care now to 

 criticize Reade's calculations in detail, for his data were in many 

 respects defective, and the conjectural element in his reasoning was 

 very large. In spite of these obvious objections, however, Reade's 

 work has shown certain statistical possibilities, and has pointed out 

 a line of investigation that may be profitably followed. Previous 

 to the appearance of Reade's memoir the mechanical sediments had 

 been used in estimating geological time, but the dissolved matter, 

 which is of nearly equal importance and much more easily measur- 

 able, was neglected. 



Professor Joly's line of attack upon the time problem was anal- 

 agous to Reade's, but different. Taking as his fundamental datum 

 the quantity of sodium in the ocean, and then estimating the annual 

 amount of sodium brought in by waters from the land, he was able 

 to compute the time required for the oceanic accumulation. This 

 method of calculation is simple, direct and clear, provided the rate 

 of supply has been constant, and that all corrections are known and 

 applied. Uncorrected, the time needed for the observed accumula- 

 tion is 99,400,000 years, which is evidently a maximum. Corrected, 

 by a questionable allowance for pre-sedimentary sodium in the 

 ocean, and for cyclic salts lifted by winds from the sea and returned 

 to it again, the estimate is reduced to 89,300,000 years. If we admit 

 that Joly's data are correct, we may round off his figures to between 

 ninety and one hundred millions of years, and feel reasonably con- 

 fident that the time of sedimentation is a quantity of that order.^ 



There are two weak points in Joly's calculation. First, the mag- 



^ For discussions of Joly's memoir see Mackie, Trans. Edin. Geol. Soc, vol. 

 8, p. 240; Fisher, Geol. Mag., 1900, p. 124; Ackroyd, Chem. News, vol. 83, p. 265; 

 84, p. 56; and Geol. Mag., Aug. and Oct. 1901. See also Sollas, "The Age of 

 the Earth," p. 21. 



