152 The American Geologist. ^^rch, i904. 
surmounted by a capital letter, represent the same formations vertically 
as do the horizontal lines across the table which follow the name of the 
formation. The figures entered in these colvimns indicate the number of 
species found common to the formation and to the other formation by 
which in the table it is vertically intersected. Where the vertical and 
horizontal columns of a single formation intersect the figures indicate 
the total number of species known from that formation. Thus in hori- 
zontal column F (Duplin Miocene), at the intersection of vertical col- 
umn J, we find the number i8, which signifies that there are eighteen 
Duplin species also found in the Qiipola (J) Oligocene. Column S is 
merely approximate, the required data to make it complete being inac- 
cessible at present. Where no common species are known the vacancy 
in the intersection indicates the fact, but it must not be inferred that 
some may not eventually be found common to the two formations. In 
the column headed 'Recent' the number of species now known to sur- 
vive in each horizon is entered. The next column shows the number of 
species now known only from that formation, numbers which future 
researches must inevitably add to. The next two columns give the Re- 
cent and Peculiar species in percentages of the total fauna of the form- 
ation to the nearest integer. In Bulletin 84 of the United States Geolog- 
ical Survey (pp. 25-31) I have discussed the question of the number of 
shell-bearing species of Mollusks which is normal to any one limited 
region. An actual count of the species found in less than one hundred 
fathoms on the coast of the United States between Cape Hatteras, North 
Carolina, and the mouth of the Savannah River aflfords the number of 
five hundred species. But in any one limited locality, as a sand beach, 
a rocky shore, an estuary or a lagoon, no one could find all these species. 
Certain kinds of terrain are more favorable to shells than others ; an 
oyster-reef or a sand beach will always have less population than a shore 
of mixed mud and gravel. So I have in the column headed 'Normal 
Fauna' given the best estimate that I could of the number of species 
which might be expected to occur in the locality where collections had 
been made to illustrate the particular formation referred to. The last 
column of the Table, headed 'Fossil faunas per cent.,' shows the percent- 
age which the forms actually found bear to the numbers theoretically 
probable. If the reasoning is correct, we can expect to find few if any 
more species than have already been collected in at least five out of the 
twelve formatinns so indicated, while of the Pleistocene of North Creek 
only about one-third of the possible species has been obtained. In a 
general way I believe these figures to be approximately true. 
"Tn the vast number of species which have been considered during the 
construction of this table it is improbable that the figures in every case 
should be mathematically exact, but it is probable that the errors if cor- 
rected would not aflfect the general conclusions drawn from the table as 
it stands. In several cases if the numbers were increased it is unlikely 
that the percentages would be materially affected. 
"The names given in the table will be familiar to the reader, but it 
seems worth while to specify that under G the name Cliesapcake applies 
