1S91.] ' ■^4:0 [Scott. 



probable relations of the atoms in the molecule, and the probability that 

 the Fe"S2 is entirely saturated. 



A structural formula for marcasite might be given as : Fe | and this 



S 

 really expresses our entire knowledge of its constitution. It may be any 

 polymer of this, for being unsaturated it should be capable of forming 

 polymers. 



I much regret that want of time has compelled me to discontinue these 

 latter investigations into the decomposibility of these minerals by solutions 

 of metallic salts under pressure, as it seems to open up a way for the study 

 of many other sulphides and would doubtless be productive of most valua- 

 ble results. Besides this it would probably adduce additional proof of the 

 correctness of my formulas for these minerals as given above. 



Acknowledgment. 



I lake this occasion to express my sense of gratitude to Prof. Edgar F. 

 Smith, who suggested the work to me, and who by his constant encourage- 

 ment and ready advice has greatly furthered its prosecution. Many of the 

 experiments were made at his suggestion, and no doubt the success of the 

 work is largely due to him. 



Notes on the Osteology of Agriocho&rus Leidy {Artionyx 0. <& W.) 



By W. B. Scott, College of New Jersey, Princeton. 



(Read before the American Philosophical Society, May 18, 1894.) 



Leidy described this genus more than forty years ago, and yet in spite 

 of the repeated explorati<jns of the White River and John Day beds, it lias 

 hitherto been known only from the skull. In 1893, Osborn and Wortman 

 published, under the name of Artionyx, an account of an extraordinary 

 liind foot, which, with atypical artiodactyl tarsus, possesses five digits and 

 very large claw-like ungual phalanges. The authors named refer 

 Artionyx to the Ancylopodaon account of the resemblance of its phalanges 

 to those of Chalicotherium. An examination of the type specimen of the 

 supposed new genus led me to believe that it represented the hind foot of 

 AgriocJicerus (see Osborn's Rise of the Mammalia in North America, p. 44, 

 separatim). This conclusion was founded upon the fact that the tarsus 

 is not only artiodactyl, but characteristically oreodont in structure, and 

 that certain features of the skull and dentition of Agriochoerus indicated 

 that it must be an exceedingly aberrant member of the oreodont group. 



Mr. J. B. Hatcher, curator of vertebrate paleontology in the Prince- 

 ton Museum, has just sent me from the White River bad lands of South 

 Dakota three fragmentary skeletons of Agriochcerus, associated with 



