SIGNIFICANCE OF WIDE DISTRIBUTION OF GRAPTOLITES 233 



thereby allowing a current to enter from the east and bring in the multi- 

 tude of graptolites. It agrees with this conclusion that the graptolite 

 horizon of the Williamson continues farthest east of all the Clinton beds 

 recognized in western New York. A similar invasion of graptolites, to 

 which we shall recur later, from the east as far west as Cincinnati had 

 already taken place in Utica time. 



Where the graptolites occur in a longer series of beds, they indicate a 

 trough or basin near an ocean. In the memoir on North American grap- 

 tolites I have pointed out the remarkable continuity of the formation of 

 graptolitiferous beds in certain regions as indicating that deposition in 

 such areas was more nearly continuous than seems to have been the case 

 in most other areas of fossiliferous rocks, or, in other words, that the 

 conditions producing the deposition of graptolite shales tended to persist 

 for a long time in the same region. From this it is inferred that long 

 series of graptolite zones indicate the former existence of long persisting 

 deep troughs in the places where these series are now found. In most, if 

 not all, cases these troughs correspond to the sites of Paleozoic geosyn- 

 clines. 



But even where the deposition of graptolite beds was apparently con- 

 tinuous for a long time, as in our Levis basin, the successive horizons are 

 not connected by transitional beds, but marked by the rather abrupt ap- 

 pearance of new forms. This means that either our knowledge is still 

 imperfect and the connecting subzones have not yet been discovered — 

 which is certainly true in some cases — or that many of them do not exist 

 in these basins. Localities like the Deep Kill, where several zones could 

 be followed bed for bed, would suggest that such transitional zones are 

 missing there. This fact, coupled with the planktonic mode of life of 

 the graptolites, indicates that this uninterrupted development has to be 

 sought in the oceanic basins, and that the horizons seen in the graptolite 

 beds are for a large part but snapshots at intervals out of this continuous 

 development in the oceans. The non-graptolitiferous intervals mark not 

 only the temporary absence of currents sweeping in from the ocean and 

 carrying the graptolites through the channel, but they probably also cover 

 intervals of non-deposition.^ 



Since the Axonolipa were fastened to seaweeds and the Axonophora 

 floated free (but judging from the relative stiffness of their axes, prob- 



2 At least Doctor Ulrich, who has given me the benefit of his extensive observations 

 in this problem with his well known liberality, writes me that he has stratigraphic evi- 

 dence that deposition in these submarginal basins was as much interrupted as that In 

 the more inland basins. 



