CONVERGENCE OF EVIDENCE 887 



presence of original lead or later alteration in the minerals by passing 

 solvents or on analytical errors. The errors in these matters for some 

 minerals have been shown to be small, under 5 per cent. For other min- 

 erals they may be several times as large. Third, the precise stratigraphic 

 position is in many cases open to considerable doubt. This may introduce 

 errors as high as 15 per cent of the total time. Fourth, the use of the 

 stratigraphic ratios to supplement the radioactive evidence introduces an- 

 other error affecting the ages of periods for which as yet no uranium 

 minerals have been found. This may be rated as high as 15 or 25 per cent, 

 but does not affect the determined points. 



Tliese several classes of errors are independent, so that to some extent 

 they tend to offset each other. The divergence of the two columns marked 

 minimum and maximum is intended to show about the latitude which 

 the summation of the probable errors gives to the results, but, as stated 

 before, the column of minimum figures should probably be given the 

 greater weight. 



CRESCENDOES IN DIASTROFHISM DUE TG COMPOSITE RHYTHMS 



The distribution of time given in this new table brings out a notable 

 regularity in the length of the periods, considering the Eocene and Oligo- 

 cene to have the value of a single older period and the Cambrian and 

 Ordovician as being equivalent to several of the later periods. The value 

 of this rhythm of the periods ranges in general from 35,000,000 to 15,- 

 000,000 years, but the extreme limits are quite uncertain. The Comanche, 

 for example, given a minimum length of 25,000,000 years, was a time of 

 but moderate diastrophism. The rate of denudation may have been lower 

 for it than for the adjacent periods, and if such were true the stratigraphic 

 measurement of its length would give an underestimate unless the lower 

 rate were considered. On the other hand, if the Mississippian, given here 

 a length of from 40,000,000 to 50,000,000 years, should be subdivided 

 into two periods, as Schuchert and Ulrich are inclined to hold, the length 

 of these would appear to be not much greater than 25,000,000 years each. 



The periods, 35,000,000 to 45,000,000 years in length, are delimited by 

 epochs of notable diastrophism of an order of magnitude which Schuchert 

 has named "disturbances.^' But the periods are subdivided by minor 

 movements leading to "breaks" in the stratigraphic record, and are com- 

 bined into larger groups constituting eras. The eras are terminated by 

 periods in which the diastrophism is recurrent and reaches the propor- 

 tions of a "revolution." The personal equation enters to a considerable 

 extent in classifying the magnitude of these division lines, and a tendency 



