30 THE MERYCOIDODONTID/E 



genera from Eporeodon, which is present, and most likely back of that from Merycoidodon, which 

 is not certainly known. From Eporeodon then was derived Promerycochoerus in one stock, Paro- 

 reodon leading into Tichole-ptus in another, Oreodontoides representing the Merychyus line, while 

 Cyclopidius is from a source unknown but probably the same as in the Great Plains. These are all 

 contemporaneous with their eastern representatives. If we do not assume this major diphyletic 

 development from some common Eocene ancestor, then we must presume that there were several 

 waves of migration, each during the same respective geologic times. The former hypothesis is pos- 

 sibly the more logical. 



There are undoubtedly more species in certain genera than are valid, in the John Day, as well as 

 in other areas, to wit, Eporeodon and Promerycochoerus, but we have several thousand feet of strata 

 in the upper Oligocene and lower Miocene and a large territory over which these forms are dissemi- 

 nated. Furthermore, in the John Day region the North Fork and Cottonwood Creek areas are 

 horizons that appear to be different from each other and from the major portion of the middle 

 and upper John Day proper. 



In conclusion, in spite of the immense amount of merycoidodont material known, there are 

 many gaps still to be filled, and a complete and proper understanding of many of the relationships 

 can come only through the recovery and intensive study of many more specimens with complete 

 skeletons. Scott, Matthew, and Loomis have done more than any other paleontologists to bring 

 order out of this maze of forms, and, while my phylogeny is somewhat different from theirs, yet 

 our differences are not fundamental. 



