STRATIGRAPIIIC TAXONOMY 593 



pelagic forms, like the graptolites and Styliolina, their invasion of conti- 

 nental basins may have been prevented by the absence of current thorough- 

 fares. In either of the latter cases we may Justly infer not only that 

 ill-recorded time intervened, but also that diastrophic movements oc- 

 curred which opened the way for freer intermigration. The introduction 

 of new types therefore is always of possibly high significance and gen- 

 erally worth searching inquir}'. And from the standpoint of stratigraphic 

 taxonomy it is always of greater significance than is the presence of sur- 

 viving older species. The new things show that something has happened, 

 and that the happening was unusual is clearly indicated also by the 

 partial extinction of the preceding fauna. 



As intimated, the validity of the principle is universally admitted in 

 "the case of proved faunal recurrences like those of the Spergen. In these 

 50 to 80 per cent of the total fauna may be recurrent. According to the 

 old method of correlation by matching of faunas, formations having 

 anything like such a percentage of species in common were classified as 

 unquestionably homotaxial or contemporaneous. But in the case of the 

 Spergen we know from positive stratigraphic relations that its fauna in- 

 vaded the Mississippi province at several widely different times. In 

 consequence the indexical value of this fauna has lost much of its exact- 

 ness. Indeed, it is practically disregarded when it comes to detailed 

 correlations, dependence being now placed on the really diagnostic species 

 which accompanied each of its several invasions. 



Early Tennessean beds included in the "Boone." — The principle beino- 

 recognized in these proved cases, why should it not apply also in those 

 other cases in which recurrence has not been established by stratigraphic 

 criteria but is as yet indicated only by the introduction of new faunal 

 elements? A case in point is the upper part of the Boone chert forma- 

 tion in southwestern Missouri. This formation, including the beds in 

 question, has hitherto been thought to be the equivalent of the Burling- 

 ton and Keokuk limestones of Iowa. So far as known, the developinent 

 of the Boone in the Yellville, Arkansas, quadrangle, where I studied the 

 section in 1902, seems quite in accord with the prevailing conception of 

 the stratigraphic position of this formation.'^ Subsequently I had occa- 

 sion to study the Boone at other localities in Arkansas and Missouri and 

 found that as mapped by geologists it frequently embraced beds at the 

 top that must be younger than the Keokuk. 



These upper beds are well developed at Gravette, Arkansas, and to the 

 north in the Joplin and Carthage districts of Missouri. They begin with 



"s Professional Paper IT. S. Geol. Survey. No. 24. 1904. p. 101. and table facing p. 90. 



