PREFACE. 



In the winter of 1892, while a student at Harvard University, I began 

 a study of the Eocene corals of the United States, and have continued it, 

 with considerable interruption, during six years. 



The study of corals presents peculiar difficulties. The classification 

 depends to a lai'ge extent upon minute and obscure structures that are fre- 

 quently destroyed in fossil specimens; and the system of classification, as 

 so many of the later students of these organisms have pointed out, is thor- 

 oughly unsatisfactory, but as yet no good system which can replace the 

 old one has been proposed, although recent investigators have accumulated 

 much valuable data that will jtrobably serve as the foundation for a better 

 and more natural system in the near future. 



Another difficulty with which I have had to contend is one that con- 

 fronts all students of sv'stematic zoology, and especially those of paleo- 

 zoology, viz. What shall constitute a species? In many instances I have 

 had enormous amounts of material from successive horizons. It may be 

 set down as a law, that the difficulty of defining species in such collections 

 varies with the size of the collection. I have tried to be conservative, 

 and when a form in one horizon grades into a form in the horizon next 

 above, I have called them varieties of the same species, even when the 

 variety possesses an individuality that makes it easily recognizable. In 

 several instances diagrams showing the phylogeny of the varieties are 

 given. These large suites of specimens have shown certain biologic facts 

 that are indisputable, yet how tliese facts should be presented may be open 

 to discus.sion. My aim has been to present them as clearly as possible, and 

 in the simplest manner, and I hiipe that my method will prove satisfactory. 



The objects of this paper should be here set forth: First, an attempt 



has been made to define every species, to present good figures of it, by 



which it can be easily identified, and to give its stratigraphic and geographic 



distribution with all precision possible. Many species of fossil corals are 



u 



