274 NEW Y0RK STATE MUSEUM 



The specimen shows the lower or nonserrated surface, and several of 

 the longer branches are turned sufficiently on one side to show the serrations 

 in a tolerable degree of perfection. 



As horizon is given the " shales of the upper part of the Hudson River 

 group." 



Assuming that the drawing is made with the care and exactitude by 

 which Professor Whitfield's graptolite drawings excel, we can infer from it 

 that the general structure of the rhabdosome is that of an Amphigraptus 

 the secondary branches being disposed in pairs so closely at the proximal 

 end that they assume a radial position ; and that they exhibit the same 

 rigidity and straightness as those of A. divergens, with the addi- 

 tional feature that the secondary branches divide again. But since the 

 simple or compound structure of the secondary branches is especially cited 

 among the diagnostic characters of the genus by the British monographers, 

 this complication beyond the genotype does not exclude the form from the 

 genus. The thecae seem also to possess, judging from the original drawing, 

 the low inclination and small amount of overlap, characteristic of the lepto- 

 graptid type of thecae. 



Hopkinson had cited this species as the second representative of his 

 genus, Clematograptus. It is, however, obvious from his figure that his 

 genotype, Clematograptus implicatus, belongs to an entirely 

 different family and is probably a dendroid form, and that hence the 

 reference of the New York species to that genus is erroneous. 



Subfamily nemagraptidae Ruedemann 



NEMAGRAPTUS EmmOIlS 



(= Coenograptus Hall) 

 We have in the first part of the Graptolites of New York [Mem. 7, 

 p. 701] introduced a form from the Deepkill shales as representing a new 

 genus and species (Sigmagraptus p r a e c u r sor) and as being ances- 

 tral to the important genus here under discussion. At the same place we 

 have proposed the new family Coenograptidae for the group of genera 

 probably derivable from Sigmagraptus. 



