886 J. BARRELL MEASUREMENTS OF GEOLOGIC TIME 



rather than on an absolute length as stated in years. Their results have 

 been brought together in the following table and the ratios are given a 

 numerical value such that unity represents 1,000,000 years as their best 

 judgment. Their results have been carried out to fractions of a million 

 years in order that the sums should be consistent, but the fractions have 

 jio other significance, since these authors have disclaimed a belief in the 

 reliability of aiiy such exact estimates of time. 



The data from the various lines of evidence are assembled in the table, 

 the conclusions being expressed in the right-hand columns. That desig- 

 juited the minimum is regarded as the more probable, but it is desirable 

 to give maximum and minimum estimates in order to prevent a single 

 column of figures conveying the idea of a precision or certainty which is 

 not yet attained. The date of the opening of the Paleozoic is the most 

 uncertain feature of this table because of the lack of uranium minerals 

 of early Paleozoic age. Certain Precambrian dates are much better deter- 

 mined. These may be listed as follows : 



Intrusion of post-Kalevian granites 925,000,000 years ago. 

 Intrusion of post-Bottnian granites 1,125,000,000 years ago. 

 Intrusion of post-Ladogian granites 1,400,000,000 years ago. 



The oldest of these rests as yet, however, on very meager data. 



The ratios of the length of the three fossil-bearing eras to each other 

 approach more nearly to the ratios given by Walcott in 1893, followed by 

 Schuchert in 1910, than to those given by others, but the absolute lengths 

 which are assigned are from 18 to 32 times greater. These absolute 

 lengths approach nearest to the estimates of Goodchild. 



In regard to the Precambrian, a difficulty is encountered in giving ap- 

 propriate divisions. It seems probable that on the same basis of crustal 

 revolutions attended by notable granitic intrusion on which the later eras 

 have been established, there may be as many as four Precambrian eras 

 following the unknown Primordial time. As no settled opinion prevails, 

 either as to the number of these eras, the best division points, or their 

 names, and as the proper correlation of widely separated regions is still 

 open to doubt, there is given here merely a list of the dates and conse- 

 quent time intervals between points determined by granitic intrusion. 

 The Fenno-Scandian nomenclature is used because the radioactive meas- 

 urements have been somewhat more fully determined with reference to it. 



As to the degree of accuracy of the table: First, the absolute ages de- 

 pend fundamentally on the determination of the radioactive constants. 

 It does not seem likely that future research will change the half-value 

 period of uranium more than 2 per cent. They depend, secondly, on the 



