248 University of California Publications in Geology [Vol. 9 



them have been brought together, we therefore are in a position to 

 sum up the evidence and work out the faunal relations which exist 

 among all of these formations. In general there are two methods 

 on which faunal correlations are based: (1) the use of character- 

 istic species; (2) the percentage method. The latter may include 

 the percentage of species common among the formations themselves 

 as well as the percentage of living species common to the formations. 

 In the present work it seems advisable to use all the methods avail- 

 able in order to obtain as nearly perfect results as possible. 



In applying the percentage method a system of tables has been 

 constructed which shows in numerical values the approximate rela- 

 tion among the formation or portions of formations. The first 

 table (table 1), shows the occurrence of all of the species that were 

 obtained from all of the formations which have been considered in 

 this investigation. Table 2 shows in numerical values the number of 

 living species found in each formation, and also the number of 

 species common to all of the formations among themselves. To 

 show this in tabulated form, all of the formations are arranged as 

 ordinates and also as abscissae. The square space opposite any two 

 of these formations, which are arranged normal to each other, con- 

 tains the number of species common to both of those formations. It 

 is evident that the number of species from the different formations 

 varies considerably, and since this is the case the magnitude of the 

 number itself can not be taken as a direct index of the relation 

 between the formations. The percentage of species common to 

 the formations compared will not vary directly as the number of 

 species. Table 3 has been constructed to show this percentage and 

 also the percentage of living species common to the formations. There 

 are two percentages to be taken into consideration ; the percentage of 

 the species in A that is common to B, and the percentage of the species 

 in B that is common to A. It is certain that these percentages will 

 usually be different, since there is, in most cases, a greater number 

 of species in one formation than in the other. For our purpose in 

 correlating, these percentages can be added and used as a factor. 

 The magnitude of this factor will indicate the approximate relation 

 between any two formations : that is, if the combined percentages of 

 the species common to A and B is greater than the combined per- 

 centages of the species common to B and C, it will indicate that A 

 and B are more closely related than B and C. 



There are a few sources of error which will affect the results 



