GRAPTOLITES OF NEW YORK, PART 2 85 



largest and most protruding rhabdosomes of the entire stock where the 

 rudder would be most effective, bear the largest disks, while the young 

 rhabdosomes, which are more or less hidden in the crowd of rhabdosomes, 

 do not possess an)-. 



Only separate rhabdosomes having been found, the question could be 

 raised whether not one and the same synrhabdosome might have borne the 

 various modifications of the appendages, characteristic of peltifer, s i g - 

 num and trident at us in such places, where they would have been 

 most effective, as for instance the tridentatus type of appendages at the 

 upper periphery of the bundle of rhabdosomes. In that case, of course, they 

 could not be considered as indicating varieties of the species but would dem- 

 onstrate a rather remarkable division of functions. The crucial test for the 

 solution of this question lies in the finding of synrhabdosomes with the vari- 

 ous appendages. Meanwhile, it may be stated that on the slabs bearing 

 the broken rhabdosomes those with like or similar appendages are so much 

 more frequently associated as to suggest that each synrhabdosome bore 

 appendages of but one kind. 



c Influence of spines on development of retioloid structure in the periderm 

 There exists, in the writer's opinion, a distinct relation, which it is 

 proper to point out in this place, between the development of the spines 

 and that of the stronger ledges of the second peridermal layer. In spine- 

 less diplograptids as D. foliaceus there is not even in strongly mace- 

 rated specimens, a trace of a system of retioloid fibers or ledges observable 

 beneath the continuous periderm. In Glossograptus quadrimu- 

 cronatus, which has rather short, paired apertural spines, a system of 

 mouth and parietal ledges is present which, however, is so weakly devel- 

 oped in comparison to the stronger main layer of the periderm that it is 

 but rarely observed. In Glossograptus ciliatus where the spines 

 are very prominent, the peridermal layer is still present, but the retioloid 

 layer has become so strong that in macerated specimens, it is found to form 

 a continuous network with four longitudinal main-ledges besides the nema- 

 caulus. The periderms of Lasiograptus mucronatus, L. bimu- 



