102 NEW YORK STATE MUSEUM 



to the section plane the thecae are always cut through giving us the appear- 

 ance shown by Diplograptus trifidus Gurley. 



We thus see that the so called "cell denticles" upon the shapes of 

 which the older writers laid so much stress as bases for specific distinction 

 have at most only a very indirect relation to the actual specific characters. 

 In other words their average constancy is merely a function of the averages 

 of direction of shale-splitting, of compression and of polypary resistance, 

 and we also see that their liability to variety of appearances must be rela- 

 tively to that of the species itself (i. e. the normal facies) very much greater, 

 being the resultant of the joint action of three variables instead of one. 



h Axes of the Dicranograptidae 

 Freeh [1897, p.615] has placed Dicranograptus (with Dicellograptus as 

 a subgenus) in his Order Axonophora and figured a specimen of Dicel- 

 lograptus divaricatus Hall var. r igi dus Lapworth [ibid. p. 620], 

 stating that this possesses an especially strong virgula, the latter being 

 figured as branching at the point of bifurcation of the branches 

 and extending into both branches. While originally accept- 

 ing the presence of such a virgula in the branches as an 

 established fact, the writer has come, from his observations, to 

 doubt the presence of any axis — either tubular as the neraa- 

 caulus or solid as the virgula — in the uniserial branches of 

 the Dicellograptidae. 



There has been observed by the writer a young specimen 



Fig. 41 D i era 11- 



sus^hLu. 5 stcuiar °f Dicranograptus ramosus, in which a nema pro- 

 portion of rhabdo- ... ... 



some with nema. x 5 trudes from the antisicular extremity of the uncomplete!.] 

 biserial portion [see text fig. 41], indicating that the latter grew along this 

 tube, by means of which the sicula was suspended from the primary disk ; 

 but there has been found no trace of any nema or virgula in any of the 

 branches in spite of continuous attention to this problem during the study 

 of a great number of specimens. Nor lias, as already stated in other 

 places, been anywhere observed the protrusion of a thread beyond the 

 broken antisicular ends of any of the uniserial branches. The direct 

 evidence of our material is hence against the assumption of the presence of 

 axes in the branches. 



